ImageApi¶

class ImageApi[source]¶

Bases: supervisely.api.module_api.RemoveableBulkModuleApi

API for working with Image. ImageApi object is immutable.

Parameters

api : Api: API connection to the server

Usage example

import os
from dotenv import load_dotenv

import supervisely as sly

# Load secrets and create API object from .env file (recommended)
# Learn more here: https://developer.supervisely.com/getting-started/basics-of-authentication
if sly.is_development():
    load_dotenv(os.path.expanduser("~/supervisely.env"))
api = sly.Api.from_env()

# Pass values into the API constructor (optional, not recommended)
# api = sly.Api(server_address="https://app.supervise.ly", token="4r47N...xaTatb")

image_info = api.image.get_info_by_id(image_id) # api usage example

Methods

`add_custom_sort`	Use this context manager to set the key name of meta object that will be used for custom sorting.
`add_tag`	Add tag with given ID to Image by ID.
`add_tag_batch`	Add tag with given ID to Images by IDs.
`check_existing_hashes`	Checks existing hashes for Images.
`check_existing_links`	Checks existing links for Images.
`check_image_uploaded`	Checks if Image has been uploaded.
`convert_info_to_json`	_convert_info_to_json
`copy`	Copies Image with given ID to destination Dataset.
`copy_batch`	Copies Images with given IDs to Dataset.
`copy_batch_optimized`	Copies Images with given IDs to Dataset.
`download`	Downloads Image from Dataset to local path by ID.
`download_blob_file`	Downloads blob file from Supervisely storage by download ID of any Image that belongs to this file.
`download_blob_file_async`	Downloads blob file from Supervisely storage by download ID asynchronously.
`download_blob_files_async`	Downloads multiple blob files from Supervisely storage by download IDs asynchronously.
`download_bytes`	Download Images with given IDs from Dataset in Binary format.
`download_bytes_generator_async`	Downloads Image bytes with given ID in batch asynchronously.
`download_bytes_many_async`	Downloads Images bytes with given IDs asynchronously and returns reults in the same order as in the input list.
`download_bytes_single_async`	Downloads Image bytes with given ID.
`download_np`	Download Image with given id in numpy format.
`download_np_async`	Downloads Image with given ID in NumPy format asynchronously.
`download_nps`	Download Images with given IDs in numpy format.
`download_nps_async`	Downloads Images with given IDs in NumPy format asynchronously.
`download_nps_by_hashes`	Download Images with given hashes in Supervisely server in numpy format.
`download_nps_by_hashes_generator`	rtype `Generator`[`Tuple`[`str`, `ndarray`], `None`, `None`]
`download_nps_generator`	rtype `Generator`[`Tuple`[`int`, `ndarray`], `None`, `None`]
`download_path`	Downloads Image from Dataset to local path by ID.
`download_path_async`	Downloads Image with given ID to local path.
`download_paths`	Download Images with given ids and saves them for the given paths.
`download_paths_async`	Download Images with given IDs and saves them to given local paths asynchronously.
`download_paths_by_hashes`	Download Images with given hashes in Supervisely server and saves them for the given paths.
`edit`	Updates the information about the image by given ID with provided parameters.
`exists`	Check if image with given name exists in dataset with given id.
`get_blob_offsets_file`	Get file with blob images offsets from the team storage.
`get_filtered_list`	List of filtered Images in the given `Dataset`.
`get_free_name`	Generates a free name for an entity with the given parent_id and name.
`get_free_names`	Returns list of free names for given dataset.
`get_info_by_id`	Get Image information by ID.
`get_info_by_id_batch`	Get Images information by ID.
`get_info_by_name`	Returns image info by image name from given dataset id.
`get_list`	List of Images in the given `Dataset`.
`get_list_all_pages`	Get list of all or limited quantity entities from the Supervisely server.
`get_list_all_pages_generator`	This generator function retrieves a list of all or a limited quantity of entities from the Supervisely server, yielding batches of entities as they are retrieved
`get_list_generator`	Returns a generator that yields lists of images in the given `Dataset` or `Project`.
`get_list_generator_async`	Yields list of images in dataset asynchronously page by page.
`get_list_idx_page_async`	Get the list of items for a given page number.
`get_list_page_generator_async`	Yields list of images in dataset asynchronously page by page.
`get_project_id`	Gets Project ID by Image ID.
`group_images_for_multiview`	Group images for multi-view by tag with given name.
`info_sequence`	Get list of all `ImageInfo` field names.
`info_tuple_name`	Get string name of `ImageInfo` NamedTuple.
`move`	Moves Image with given ID to destination Dataset.
`move_batch`	Moves Images with given IDs to Dataset.
`move_batch_optimized`	Moves Images with given IDs to Dataset.
`preview_url`	Previews Image with the given resolution parameters.
`raise_name_intersections_if_exist`	Raises error if images with given names already exist in dataset.
`remove`	Remove image from supervisely by id.
`remove_batch`	Remove images from supervisely by IDs.
`rename`	Renames Image with given ID.
`set_custom_sort`	Sets custom sort value for image with given ID.
`set_custom_sort_bulk`	Sets custom sort values for images with given IDs.
`set_remote`	This method helps to change local source to remote for images without re-uploading them as new.
`storage_url`	Get full Image URL link in Supervisely server.
`update_custom_sort`	Updates a copy of the meta dictionary with a new custom sort value.
`update_meta`	It is possible to add custom JSON data to every image for storing some additional information.
`update_tag_value`	Update tag value with given ID.
`upload_blob_images`	Uploads images from blob file in Team Files to dataset.
`upload_by_offsets`	Upload images from blob file in Team Files by offsets to Dataset with prepared names.
`upload_by_offsets_generator`	Upload images from blob file in Team Files by offsets to Dataset.
`upload_dir`	Uploads all images with supported extensions from given directory to Supervisely.
`upload_dirs`	Uploads all images with supported extensions from given directories to Supervisely.
`upload_hash`	Upload Image from given hash to Dataset.
`upload_hashes`	Upload images from given hashes to Dataset.
`upload_id`	Upload Image by ID to Dataset.
`upload_ids`	Upload Images by IDs to Dataset.
`upload_link`	Uploads Image from given link to Dataset.
`upload_links`	Uploads Images from given links to Dataset.
`upload_medical_images`	Upload medical 2D images (DICOM) to Supervisely and group them by specified or default tag.
`upload_multispectral`	Uploads multispectral image to Supervisely, if channels are provided, they will be uploaded as separate images.
`upload_multiview_images`	Uploads images to Supervisely and adds a tag to them.
`upload_np`	Upload given Image in numpy format with given name to Dataset.
`upload_nps`	Upload given Images in numpy format with given names to Dataset.
`upload_path`	Uploads Image with given name from given local path to Dataset.
`upload_paths`	Uploads Images with given names from given local path to Dataset.
`url`	Gets Image URL by ID.

Attributes

`MAX_WAIT_ATTEMPTS`	Maximum number of attempts that will be made to wait for a certain condition to be met.
`WAIT_ATTEMPT_TIMEOUT_SEC`	Number of seconds for intervals between attempts.

InfoType¶: alias of supervisely.api.module_api.ImageInfo

add_custom_sort(key)[source]¶

Use this context manager to set the key name of meta object that will be used for custom sorting. This context manager allows you to set the sort_by attribute of ImageApi object for the duration of the context, then delete it. If nested functions support this functionality, each image they process will automatically receive a custom sorting parameter based on the available meta object.

Parameters

key : str: It is a key name of meta object that will be used for sorting.

add_tag(image_id, tag_id, value=None)[source]¶

Add tag with given ID to Image by ID.

Parameters

image_id : int: Image ID in Supervisely.
tag_id : int: Tag ID in Supervisely.
value : int or str or None, optional: Tag value.

Returns

None

Return type

NoneType

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

image_id = 2389126
tag_id = 277083
api.image.add_tag(image_id, tag_id)

add_tag_batch(image_ids, tag_id, value=None, progress_cb=None, batch_size=100, tag_meta=None)[source]¶

Add tag with given ID to Images by IDs.

Parameters

image_ids : List[int]: List of Images IDs in Supervisely.
tag_id : int: Tag ID in Supervisely.
value : int or str or None, optional: Tag value.
progress_cb : tqdm or callable, optional: Function for tracking progress of adding tag.
batch_size : int, optional: Batch size
tag_meta : TagMeta, optional: Tag Meta. Needed for value validation, omit to skip validation

Returns

None

Return type

NoneType

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

image_ids = [2389126, 2389127]
tag_id = 277083
api.image.add_tag_batch(image_ids, tag_id)

check_existing_hashes(hashes, progress_cb=None)[source]¶

Checks existing hashes for Images.

Parameters

hashes : List[str]: List of hashes.
progress_cb : tqdm or callable, optional: Function for tracking progress of checking.

Returns

List of existing hashes

Return type

List[str]

Usage example

Checkout detailed example here (you must be logged into your Supervisely account)

# Helpful method when your uploading was interrupted
# You can check what images has been successfully uploaded by their hashes and what not
# And continue uploading the rest of the images from that point

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

# Find project
project = api.project.get_info_by_id(WORKSPACE_ID, PROJECT_ID)

# Get paths of all images in a directory
images_paths = sly.fs.list_files('images_to_upload')

# Calculate hashes for all images paths
hash_to_image = {}
images_hashes = []

for idx, item in enumerate(images_paths):
    item_hash = sly.fs.get_file_hash(item)
    images_hashes.append(item_hash)
    hash_to_image[item_hash] = item

# Get hashes that are already on server
remote_hashes = api.image.check_existing_hashes(images_hashes)
already_uploaded_images = {hh: hash_to_image[hh] for hh in remote_hashes}

check_existing_links(links, progress_cb=None, team_id=None)[source]¶

Checks existing links for Images.

Parameters

links : List[str]: List of links.
progress_cb : tqdm or callable, optional: Function for tracking progress of checking.
team_id : int: Team ID in Supervisely (will be used to get remote storage settings).

Returns

List of existing links

Return type

List[str]

check_image_uploaded(hash)[source]¶

Checks if Image has been uploaded.

Parameters

hash : str: Image hash in Supervisely.

Returns

True if Image with given hash exist, otherwise False

Return type

bool

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

image_check_uploaded = api.image.check_image_uploaded("YZKQrZH5C0rBvGGA3p7hjWahz3/pV09u5m30Bz8GeYs=")
print(image_check_uploaded)
# Output: True

classmethod convert_info_to_json(info)¶

_convert_info_to_json

Return type: Dict

copy(dst_dataset_id, id, change_name_if_conflict=False, with_annotations=False)[source]¶

Copies Image with given ID to destination Dataset.

Parameters

dst_dataset_id : int: Destination Dataset ID in Supervisely.
id : int: Image ID in Supervisely.
change_name_if_conflict : bool, optional: If True adds suffix to the end of Image name when Dataset already contains an Image with identical name, If False and images with the identical names already exist in Dataset raises error.
with_annotations : bool, optional: If True Image will be copied to Dataset with annotations, otherwise only Images without annotations.

Returns

Information about Image. See info_sequence

Return type

ImageInfo

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

dst_ds_id = 365184
img_id = 121236920

img_info = api.image.copy(dst_ds_id, img_id, with_annotations=True)

copy_batch(dst_dataset_id, ids, change_name_if_conflict=False, with_annotations=False, progress_cb=None)[source]¶

Copies Images with given IDs to Dataset.

Parameters

dst_dataset_id : int: Destination Dataset ID in Supervisely.
ids : List[int]: Images IDs in Supervisely.
change_name_if_conflict : bool, optional: If True adds suffix to the end of Image name when Dataset already contains an Image with identical name, If False and images with the identical names already exist in Dataset raises error.
with_annotations : bool, optional: If True Image will be copied to Dataset with annotations, otherwise only Images without annotations.
progress_cb : tqdm or callable, optional: Function for tracking the progress of copying.

Raises

TypeError if type of ids is not list

Raises

ValueError if images ids are from the destination Dataset

Returns

List with information about Images. See info_sequence

Return type

List[ImageInfo]

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

ds_lemon_id = 1780

ds_lemon_img_infos = api.image.get_list(ds_lemon_id)

lemons_img_ids = [lemon_img_info.id for lemon_img_info in ds_lemon_img_infos]

ds_fruit_id = 2574
ds_fruit_img_infos = api.image.copy_batch(ds_fruit_id, lemons_img_ids, with_annotations=True)

copy_batch_optimized(src_dataset_id, src_image_infos, dst_dataset_id, with_annotations=True, progress_cb=None, dst_names=None, batch_size=500, skip_validation=False, save_source_date=True)[source]¶

Copies Images with given IDs to Dataset.

Parameters

src_dataset_id : int: Source Dataset ID in Supervisely.
src_image_infos : List [ ImageInfo ]: ImageInfo objects of images to copy.
dst_dataset_id : int: Destination Dataset ID in Supervisely.
with_annotations : bool, optional: If True Image will be copied to Dataset with annotations, otherwise only Images without annotations.
progress_cb : tqdm or callable, optional: Function for tracking the progress of copying.
dst_names : List [ ImageInfo ], optional: ImageInfo list with existing items in destination dataset.
batch_size : int, optional: Number of elements to copy for each request.
skip_validation : bool, optional: Flag for skipping additinal validations.
save_source_date : bool, optional: Save source annotation dates (creation and modification) or create a new date.

Raises

TypeError if type of src_image_infos is not list

Returns

List with information about Images. See info_sequence

Return type

List[ImageInfo]

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

src_ds_id = 2231
img_infos = api.image.get_list(src_ds_id)

dest_ds_id = 2574
dest_img_infos = api.image.copy_batch_optimized(src_ds_id, img_infos, dest_ds_id)

download(id, path)[source]¶

Downloads Image from Dataset to local path by ID.

Parameters

id : int: Image ID in Supervisely.
path : str: Local save path for Image.

Returns

None

Return type

NoneType

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

img_info = api.image.get_info_by_id(770918)
save_path = os.path.join("/home/admin/work/projects/lemons_annotated/ds1/test_imgs/", img_info.name)

api.image.download_path(770918, save_path)

download_blob_file(project_id, download_id, path=None, log_progress=True, chunk_size=None)[source]¶

Downloads blob file from Supervisely storage by download ID of any Image that belongs to this file.

Parameters

project_id : int: Project ID in Supervisely.
download_id : str: Download ID of any Image that belongs to the blob file in Supervisely storage.
path : str, optional: Path to save the blob file. If None, returns blob file content as bytes.
progress_cb : tqdm or callable, optional: Function for tracking download progress.
chunk_size : int, optional: Size of chunk for streaming. Default is 8 MB.

Returns

Blob file content if path is None, otherwise None.

Return type

bytes or None

Usage example

api = sly.Api.from_env()


image_id = 6789
image_info = api.image.get_info_by_id(image_id)
project_id = api.dataset.get_info_by_id(image_info.dataset_id).project_id

# Download and save to file
api.image.download_blob_file(project_id, image_info.download_id, "/path/to/save/archive.tar")

# Get archive as bytes
archive_bytes = api.image.download_blob_file(project_id, image_info.download_id)

async download_blob_file_async(project_id, download_id, path, semaphore=None, log_progress=True, progress_cb=None)[source]¶

Downloads blob file from Supervisely storage by download ID asynchronously.

Parameters

project_id : int: Project ID in Supervisely.
download_id : str: Download ID of any Image that belongs to the blob file in Supervisely storage.
path : str: Path to save the blob file.
semaphore : asyncio.Semaphore, optional: Semaphore for limiting the number of simultaneous downloads.
log_progress : bool, optional: If True, shows progress bar.
progress_cb : tqdm or callable, optional: Function for tracking download progress.

async download_blob_files_async(project_id, download_ids, paths, semaphore=None, log_progress=True, progress_cb=None)[source]¶

Downloads multiple blob files from Supervisely storage by download IDs asynchronously.

Parameters

project_id : int: Project ID in Supervisely.
download_ids : List[str]: List of download IDs of any Image that belongs to the blob files in Supervisely storage.
paths : List[str]: List of paths to save the blob files.
semaphore : asyncio.Semaphore, optional: Semaphore for limiting the number of simultaneous downloads.
log_progress : bool, optional: If True, shows progress bar.
progress_cb : tqdm or callable, optional: Function for tracking download progress.

download_bytes(dataset_id, ids, progress_cb=None)[source]¶

Download Images with given IDs from Dataset in Binary format.

Parameters

dataset_id : int: Dataset ID in Supervisely, where Images are located.
ids : List[int]: List of Image IDs in Supervisely.
progress_cb : tqdm or callable, optional: Function for tracking download progress.

Returns

List of Images in binary format

Return type

List[bytes]

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

img_bytes = api.image.download_bytes(dataset_id, [770918])
print(img_bytes)
# Output: [b'ÿØÿàJFIF\ ...']

async download_bytes_generator_async(dataset_id, img_ids, semaphore=None, headers=None, check_hash=False, progress_cb=None, progress_cb_type='number')[source]¶

Downloads Image bytes with given ID in batch asynchronously. Yields tuple of Image ID and bytes of downloaded image. Uses bulk download API method.

Parameters

dataset_id : int: Dataset ID in Supervisely.
img_ids : List[int]: List of Image IDs in Supervisely.
semaphore : asyncio.Semaphore, optional: Semaphore for limiting the number of simultaneous downloads.
headers : dict, optional: Headers for request.
check_hash : bool, optional: If True, checks hash of downloaded bytes. Default is False.
progress_cb : Optional[Union[tqdm, Callable]]: Function for tracking download progress.
progress_cb_type : Literal["number", "size"], optional: Type of progress callback. Can be “number” or “size”. Default is “number”.

Returns

Tuple of Image ID and bytes of downloaded image.

Return type

Tuple[int, bytes]

Usage example

import supervisely as sly
import asyncio

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

dataset_id = 123456
img_ids = [770918, 770919, 770920, 770921, ... , 770992]
tasks = []
for batch in batched(img_ids, 50):
    task = api.image.download_bytes_batch_async(dataset_id, batch)
    tasks.append(task)
results = await asyncio.gather(*tasks)

async download_bytes_many_async(ids, semaphore=None, headers=None, check_hash=True, progress_cb=None, progress_cb_type='number')[source]¶

Downloads Images bytes with given IDs asynchronously and returns reults in the same order as in the input list.

Parameters

ids : List[int]: List of Image IDs in Supervisely.
semaphore : asyncio.Semaphore, optional: Semaphore for limiting the number of simultaneous downloads.
headers : dict, optional: Headers for every request.
check_hash : bool, optional: If True, checks hash of downloaded images.
progress_cb : Optional[Union[tqdm, Callable]]: Function for tracking download progress.
progress_cb_type : Literal["number", "size"], optional: Type of progress callback. Can be “number” or “size”. Default is “number”.

Returns

List of bytes of downloaded images.

Return type

List[bytes]

Usage example

import supervisely as sly
import asyncio

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN
api = sly.Api.from_env()

loop = sly.utils.get_or_create_event_loop()
semaphore = asyncio.Semaphore(100)
img_bytes_list = loop.run_until_complete(api.image.download_bytes_imgs_async(ids, semaphore))

async download_bytes_single_async(id, semaphore=None, range_start=None, range_end=None, headers=None, check_hash=True, progress_cb=None, progress_cb_type='number')[source]¶

Downloads Image bytes with given ID.

Parameters

id : int: Image ID in Supervisely.
semaphore : asyncio.Semaphore, optional: Semaphore for limiting the number of simultaneous downloads.
range_start : int, optional: Start byte of range for partial download.
range_end : int, optional: End byte of range for partial download.
headers : dict, optional: Headers for request.
check_hash : bool, optional: If True, checks hash of downloaded bytes. Check is not supported for partial downloads. When range is set, hash check is disabled.
progress_cb : Optional[Union[tqdm, Callable]]: Function for tracking download progress.
progress_cb_type : Literal["number", "size"], optional: Type of progress callback. Can be “number” or “size”. Default is “number”.

Returns

Bytes of downloaded image.

Return type

bytes

Usage example

import supervisely as sly
import asyncio

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

img_id = 770918
loop = sly.utils.get_or_create_event_loop()
img_bytes = loop.run_until_complete(api.image.download_bytes_async(img_id))

download_np(id, keep_alpha=False)[source]¶

Download Image with given id in numpy format.

Parameters

id : int: Image ID in Supervisely.
keep_alpha : bool, optional: If True keeps alpha mask for image, otherwise don’t.

Returns

Image in RGB numpy matrix format

Return type

np.ndarray

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

image_np = api.image.download_np(770918)

async download_np_async(id, semaphore=None, keep_alpha=False, progress_cb=None, progress_cb_type='number')[source]¶

Downloads Image with given ID in NumPy format asynchronously.

Parameters

id : int: Image ID in Supervisely.
semaphore : asyncio.Semaphore, optional: Semaphore for limiting the number of simultaneous downloads.
keep_alpha : bool, optional: If True keeps alpha mask for image, otherwise don’t.
progress_cb : tqdm or callable, optional: Function for tracking download progress.
progress_cb_type : Literal["number", "size"], optional: Type of progress callback. Can be “number” or “size”. Default is “number”.

Returns

Image in RGB numpy matrix format

Return type

np.ndarray

Usage example

import supervisely as sly
import asyncio
from tqdm.asyncio import tqdm

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

DATASET_ID = 98357
semaphore = asyncio.Semaphore(100)
images = api.image.get_list(DATASET_ID)
tasks = []
pbar = tqdm(total=len(images), desc="Downloading images", unit="image")
for image in images:
    task = api.image.download_np_async(image.id, semaphore, progress_cb=pbar)
    tasks.append(task)
results = await asyncio.gather(*tasks)

download_nps(dataset_id, ids, progress_cb=None, keep_alpha=False)[source]¶

Download Images with given IDs in numpy format.

Parameters

dataset_id : int: Dataset ID in Supervisely, where Images are located.
ids : List[int]: List of Images IDs in Supervisely.
progress_cb : tqdm or callable, optional: Function for tracking download progress.
keep_alpha : bool, optional: If True keeps alpha mask for Image, otherwise don’t.

Returns

List of Images in RGB numpy matrix format

Return type

List[np.ndarray]

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

image_ids = [770918, 770919, 770920]
image_nps = api.image.download_nps(dataset_id, image_ids)

async download_nps_async(ids, semaphore=None, keep_alpha=False, progress_cb=None, progress_cb_type='number')[source]¶

Downloads Images with given IDs in NumPy format asynchronously.

Parameters

ids : List[int]: List of Image IDs in Supervisely.
semaphore : asyncio.Semaphore, optional: Semaphore for limiting the number of simultaneous downloads.
keep_alpha : bool, optional: If True keeps alpha mask for images, otherwise don’t.
progress_cb : tqdm or callable, optional: Function for tracking download progress.
progress_cb_type : Literal["number", "size"], optional: Type of progress callback. Can be “number” or “size”. Default is “number”.

Returns

List of Images in RGB numpy matrix format

Return type

List[np.ndarray]

Usage example

import supervisely as sly
import asyncio

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

DATASET_ID = 98357
semaphore = asyncio.Semaphore(100)
images = api.image.get_list(DATASET_ID)
img_ids = [image.id for image in images]
loop = sly.utils.get_or_create_event_loop()
results = loop.run_until_complete(
                api.image.download_nps_async(img_ids, semaphore)
            )

download_nps_by_hashes(hashes, keep_alpha=False, progress_cb=None)[source]¶

Download Images with given hashes in Supervisely server in numpy format.

Parameters

hashes : List[str]: List of images hashes in Supervisely.
progress_cb : tqdm or callable, optional: Function for tracking download progress.

Returns

List of images

Return type

class: List[np.ndarray]

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

image_ids = [770918, 770919, 770920]
image_hashes = []

for img_id in image_ids:
    img_info = api.image.get_info_by_id(image_id)
    image_hashes.append(img_info.hash)

image_nps = api.image.download_nps_by_hashes(image_hashes)

download_path(id, path)[source]¶

Downloads Image from Dataset to local path by ID.

Parameters

id : int: Image ID in Supervisely.
path : str: Local save path for Image.

Returns

None

Return type

NoneType

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

img_info = api.image.get_info_by_id(770918)
save_path = os.path.join("/home/admin/work/projects/lemons_annotated/ds1/test_imgs/", img_info.name)

api.image.download_path(770918, save_path)

async download_path_async(id, path, semaphore=None, range_start=None, range_end=None, headers=None, check_hash=True, progress_cb=None, progress_cb_type='number')[source]¶

Downloads Image with given ID to local path.

Parameters

id : int: Image ID in Supervisely.
path : str: Local save path for Image.
semaphore : asyncio.Semaphore, optional: Semaphore for limiting the number of simultaneous downloads.
range_start : int, optional: Start byte of range for partial download.
range_end : int, optional: End byte of range for partial download.
headers : dict, optional: Headers for request.
check_hash : bool, optional: If True, checks hash of downloaded file. Check is not supported for partial downloads. When range is set, hash check is disabled.
progress_cb : tqdm or callable, optional: Function for tracking download progress.
progress_cb_type : Literal["number", "size"], optional: Type of progress callback. Can be “number” or “size”. Default is “number”.

Returns

None

Return type

NoneType

Usage example

import supervisely as sly
import asyncio

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

img_info = api.image.get_info_by_id(770918)
save_path = os.path.join("/path/to/save/", img_info.name)

semaphore = asyncio.Semaphore(100)
loop = sly.utils.get_or_create_event_loop()
loop.run_until_complete(
        api.image.download_path_async(img_info.id, save_path, semaphore)
    )

download_paths(dataset_id, ids, paths, progress_cb=None)[source]¶

Download Images with given ids and saves them for the given paths.

Parameters

dataset_id : int: Dataset ID in Supervisely, where Images are located.
ids : List[int]: List of Image IDs in Supervisely.
paths : List[str]: Local save paths for Images.
progress_cb : tqdm or callable, optional: Function for tracking download progress.

Raises

ValueError if len(ids) != len(paths)

Returns

None

Return type

NoneType

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

local_save_dir = "/home/admin/work/projects/lemons_annotated/ds1/test_imgs"
save_paths = []
image_ids = [771755, 771756, 771757, 771758, 771759, 771760]
img_infos = api.image.get_info_by_id_batch(image_ids)

p = tqdm(desc="Images downloaded: ", total=len(img_infos))
for img_info in img_infos:
    save_paths.append(os.path.join(local_save_dir, img_info.name))

api.image.download_paths(2573, image_ids, save_paths, progress_cb=p)
# Progress:
# {"message": "progress", "event_type": "EventType.PROGRESS", "subtask": "Images downloaded: ", "current": 0, "total": 6, "timestamp": "2021-03-15T19:47:15.406Z", "level": "info"}
# {"message": "progress", "event_type": "EventType.PROGRESS", "subtask": "Images downloaded: ", "current": 1, "total": 6, "timestamp": "2021-03-15T19:47:16.366Z", "level": "info"}
# {"message": "progress", "event_type": "EventType.PROGRESS", "subtask": "Images downloaded: ", "current": 2, "total": 6, "timestamp": "2021-03-15T19:47:16.367Z", "level": "info"}
# {"message": "progress", "event_type": "EventType.PROGRESS", "subtask": "Images downloaded: ", "current": 3, "total": 6, "timestamp": "2021-03-15T19:47:16.367Z", "level": "info"}
# {"message": "progress", "event_type": "EventType.PROGRESS", "subtask": "Images downloaded: ", "current": 4, "total": 6, "timestamp": "2021-03-15T19:47:16.367Z", "level": "info"}
# {"message": "progress", "event_type": "EventType.PROGRESS", "subtask": "Images downloaded: ", "current": 5, "total": 6, "timestamp": "2021-03-15T19:47:16.368Z", "level": "info"}
# {"message": "progress", "event_type": "EventType.PROGRESS", "subtask": "Images downloaded: ", "current": 6, "total": 6, "timestamp": "2021-03-15T19:47:16.368Z", "level": "info"}

async download_paths_async(ids, paths, semaphore=None, headers=None, check_hash=True, progress_cb=None, progress_cb_type='number')[source]¶

Download Images with given IDs and saves them to given local paths asynchronously.

Parameters

ids : List[int]: List of Image IDs in Supervisely.
paths : List[str]: Local save paths for Images.
semaphore : asyncio.Semaphore, optional: Semaphore for limiting the number of simultaneous downloads.
headers : dict, optional: Headers for request.
check_hash : bool, optional: If True, checks hash of downloaded images.
progress_cb : tqdm or callable, optional: Function for tracking download progress.
progress_cb_type : Literal["number", "size"], optional: Type of progress callback. Can be “number” or “size”. Default is “number”.

Raises

ValueError if len(ids) != len(paths)

Returns

None

Return type

NoneType

Usage example

import supervisely as sly
import asyncio

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

ids = [770918, 770919]
paths = ["/path/to/save/image1.png", "/path/to/save/image2.png"]
loop = sly.utils.get_or_create_event_loop()
loop.run_until_complete(api.image.download_paths_async(ids, paths))

download_paths_by_hashes(hashes, paths, progress_cb=None)[source]¶

Download Images with given hashes in Supervisely server and saves them for the given paths.

Parameters

hashes : List[str]: List of images hashes in Supervisely.
paths : List[str]: List of paths to save images.
progress_cb : tqdm or callable, optional: Function for tracking download progress.

Raises

ValueError if len(hashes) != len(paths)

Returns

None

Return type

NoneType

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

dataset_id = 447130
dir_for_save = '/home/admin/Downloads/img'
hashes = []
paths = []
imgs_info = api.image.get_list(dataset_id)
for im_info in imgs_info:
    hashes.append(im_info.hash)
    # It is necessary to save images with the same names(extentions) as on the server
    paths.append(os.path.join(dir_for_save, im_info.name))
api.image.download_paths_by_hashes(hashes, paths)

edit(id, name=None, description=None, meta=None, return_json=False)[source]¶

Updates the information about the image by given ID with provided parameters. At least one parameter must be set, otherwise ValueError will be raised.

Parameters

id : int: Image ID in Supervisely.
name : str, optional: New Image name.
description : str, optional: New Image description.
meta : dict, optional: New Image metadata. Custom additional image info that contain image technical and/or user-generated data.

Return_json

If True, return response in JSON format, otherwise convert it ImageInfo object. This parameter is only added for backward compatibility for update_meta method. It’s not recommended to use it in new code.

Raises

ValueError if at least one parameter is not set

Raises

:class:`ValueError if meta parameter was set and it is not a dictionary

Returns

Information about updated image as ImageInfo object or as dict if return_json is True

Return type

ImageInfo or dict

Usage example

import supervisely as sly

api = sly.Api.from_env()

image_id = 123456
new_image_name = "IMG_3333_new.jpg"

api.image.edit(id=image_id, name=new_image_name)

exists(parent_id, name)[source]¶

Check if image with given name exists in dataset with given id.

Parameters

parent_id : int: Dataset ID in Supervisely.
name : str: Image name in Supervisely.

Returns

True if image exists, False otherwise.

Return type

bool

get_blob_offsets_file(team_file_id, progress_cb=None)[source]¶

Get file with blob images offsets from the team storage.

Parameters

team_file_id : int: ID of the binary file in the team storage.
progress_cb : tqdm or callable, optional: Function for tracking the progress of downloading.

Returns

Path to the file with blob images offsets in temporary directory.

Return type

str

get_filtered_list(dataset_id=None, filters=None, sort='id', sort_order='asc', force_metadata_for_links=True, limit=None, return_first_response=False, project_id=None)[source]¶

List of filtered Images in the given Dataset. Differs in a more flexible filter format from the get_list() method.

Parameters

dataset_id : int: Dataset ID in which the Images are located.
filters : List[Dict], optional: List of params to sort output Images.
sort : str, optional: Field name to sort. One of {‘id’ (default), ‘name’, ‘description’, ‘labelsCount’, ‘createdAt’, ‘updatedAt’, ‘customSort’}.
sort_order : str, optional: Sort order. One of {‘asc’ (default), ‘desc’}
project_id : int: Project ID in which the Images are located.

Returns

Objects with image information from Supervisely.

Return type

List[ImageInfo]

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

# Get list of Images with names containing subsequence '2008'
img_infos = api.image.get_filtered_list(dataset_id, filters=[{ 'type': 'images_filename', 'data': { 'value': '2008' } }])

get_free_name(parent_id, name)¶

Generates a free name for an entity with the given parent_id and name. Adds an increasing suffix to original name until a unique name is found.

Parameters

parent_id : int: ID of the parent entity.
name : str: Name of the entity.

Returns

Returns free name.

Return type

str

Usage example

import supervisely as sly

# You can connect to API directly
address = 'https://app.supervise.ly/'
token = 'Your Supervisely API Token'
api = sly.Api(address, token)

# Or you can use API from environment
os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()


name = "IMG_0315.jpeg"
dataset_id = 55832
free_name = api.image.get_free_name(dataset_id, name)
print(free_name) # IMG_0315_001.jpeg

get_free_names(dataset_id, names)[source]¶

Returns list of free names for given dataset.

Parameters

dataset_id : int: Dataset ID in Supervisely.
names : List[str]: List of names to check.

Returns

List of free names.

Return type

List[str]

get_info_by_id(id, force_metadata_for_links=True)[source]¶

Get Image information by ID.

Parameters

id : int: Image ID in Supervisely.

Returns

Object with image information from Supervisely.

Return type

ImageInfo

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

# You can get Image ID by listing all images in the Dataset as shown in get_list
# Or you can open certain image in Supervisely Annotation Tool UI and get last digits of the URL
img_info = api.image.get_info_by_id(770918)

get_info_by_id_batch(ids, progress_cb=None, force_metadata_for_links=True)[source]¶

Get Images information by ID.

Parameters

ids : List[int]: Images IDs in Supervisely.
progress_cb : tqdm or callable, optional: Function for tracking the progress.

Returns

Objects with image information from Supervisely.

Return type

List[ImageInfo]

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

img_ids = [376728, 376729, 376730, 376731, 376732, 376733]
img_infos = image.get_info_by_id_batch(img_ids)

get_info_by_name(dataset_id, name, force_metadata_for_links=True)[source]¶

Returns image info by image name from given dataset id.

Parameters

dataset_id : int: Dataset ID in Supervisely, where Image is located.
name : str: Image name in Supervisely.
force_metadata_for_links : bool, optional: If True, returns full_storage_url and path_original fields in ImageInfo.

Returns

Object with image information from Supervisely.

Return type

ImageInfo

get_list(dataset_id=None, filters=None, sort='id', sort_order='asc', limit=None, force_metadata_for_links=True, return_first_response=False, project_id=None, only_labelled=False, fields=None, recursive=False)[source]¶

List of Images in the given Dataset.

Parameters

dataset_id : int: Dataset ID in which the Images are located.
filters : List[Dict], optional: List of params to sort output Images.
sort : str, optional: Field name to sort. One of {‘id’ (default), ‘name’, ‘description’, ‘labelsCount’, ‘createdAt’, ‘updatedAt’, customSort}
sort_order : str, optional: Sort order. One of {‘asc’ (default), ‘desc’}
limit : int, optional: Max number of list elements. No limit if None (default).
force_metadata_for_links : bool, optional: If True, updates meta for images with remote storage links when listing.
return_first_response : bool, optional: If True, returns first response without waiting for all pages.
project_id : int: Project ID in which the Images are located.
only_labelled : bool, optional: If True, returns only images with labels.
fields : List[str], optional: List of fields to return. If None, returns all fields.
recursive : bool, optional: If True, returns all images from dataset recursively (including images in nested datasets).

Returns

Objects with image information from Supervisely.

Return type

List[ImageInfo]

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

# Get list of Images with width = 1067
img_infos = api.image.get_list(dataset_id, filters=[{ 'field': 'width', 'operator': '=', 'value': '1067' }])
print(img_infos)
# Output: [ImageInfo(id=770915,
#                    name='IMG_3861.jpeg',
#                    link=None,
#                    hash='ZdpMD+ZMJx0R8BgsCzJcqM7qP4M8f1AEtoYc87xZmyQ=',
#                    mime='image/jpeg',
#                    ext='jpeg',
#                    size=148388,
#                    width=1067,
#                    height=800,
#                    labels_count=4,
#                    dataset_id=2532,
#                    created_at='2021-03-02T10:04:33.973Z',
#                    updated_at='2021-03-02T10:04:33.973Z',
#                    meta={},
#                    path_original='/h5un6l2bnaz1vj8a9qgms4-public/images/original/7/h/Vo/...jpg',
#                    full_storage_url='http://app.supervise.ly/h5un6l2bnaz1vj8a9qgms4-public/images/original/7/h/Vo/...jpg'),
#                    tags=[],
# ImageInfo(id=770916,
#           name='IMG_1836.jpeg',
#           link=None,
#           hash='YZKQrZH5C0rBvGGA3p7hjWahz3/pV09u5m30Bz8GeYs=',
#           mime='image/jpeg',
#           ext='jpeg',
#           size=140222,
#           width=1067,
#           height=800,
#           labels_count=3,
#           dataset_id=2532,
#           created_at='2021-03-02T10:04:33.973Z',
#           updated_at='2021-03-02T10:04:33.973Z',
#           meta={},
#           path_original='/h5un6l2bnaz1vj8a9qgms4-public/images/original/C/Y/Hq/...jpg',
#           full_storage_url='http://app.supervise.ly/h5un6l2bnaz1vj8a9qgms4-public/images/original/C/Y/Hq/...jpg'),
#           tags=[]
# ]

get_list_all_pages(method, data, progress_cb=None, convert_json_info_cb=None, limit=None, return_first_response=False)¶

Get list of all or limited quantity entities from the Supervisely server.

Parameters

method : str: Request method name
data : dict: Dictionary with request body info
progress_cb : Progress, optional: Function for tracking download progress.
convert_json_info_cb : Callable, optional: Function for convert json info
limit : int, optional: Number of entity to retrieve
return_first_response : bool, optional: Specify if return first response

get_list_all_pages_generator(method, data, progress_cb=None, convert_json_info_cb=None, limit=None, return_first_response=False)¶

This generator function retrieves a list of all or a limited quantity of entities from the Supervisely server, yielding batches of entities as they are retrieved

Parameters

method : str: Request method name
data : dict: Dictionary with request body info
progress_cb : Progress, optional: Function for tracking download progress.
convert_json_info_cb : Callable, optional: Function for convert json info
limit : int, optional: Number of entity to retrieve
return_first_response : bool, optional: Specify if return first response

get_list_generator(dataset_id=None, filters=None, sort='id', sort_order='asc', limit=None, force_metadata_for_links=False, batch_size=None, project_id=None)[source]¶

Returns a generator that yields lists of images in the given Dataset or Project.

Parameters

dataset_id : int: Dataset ID in which the Images are located.
filters : List[Dict], optional: List of params to sort output Images.
sort : str, optional: Field name to sort. One of {‘id’ (default), ‘name’, ‘description’, ‘labelsCount’, ‘createdAt’, ‘updatedAt’, customSort}
sort_order : str, optional: Sort order. One of {‘asc’ (default), ‘desc’}
limit : int, optional: Max number of list elements. No limit if None (default).
force_metadata_for_links : bool, optional: If True, updates meta for images with remote storage links when listing.
batch_size : int, optional: Number of images to get in each request.
project_id : int: Project ID in which the Images are located.

Return type

Iterator[List[ImageInfo]]

Returns

Generator that yields lists of images in the given Dataset or Project.

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

for images_batch in api.image.get_list_generator(dataset_id):
    print(images_batch)

async get_list_generator_async(dataset_id=None, filters=None, sort='id', sort_order='asc', force_metadata_for_links=True, only_labelled=False, fields=None, per_page=None, semaphore=None, **kwargs)[source]¶

Yields list of images in dataset asynchronously page by page.

Parameters

dataset_id : int: Dataset ID in Supervisely.
filters : List[Dict[str, str]], optional: Filters for images.
sort : str, optional: Field name to sort. One of {‘id’ (default), ‘name’, ‘description’, ‘labelsCount’, ‘createdAt’, ‘updatedAt’, ‘customSort’}.
sort_order : str, optional: Sort order for images. One of {‘asc’ (default), ‘desc’}
force_metadata_for_links : bool, optional: If True, forces metadata for links.
only_labelled : bool, optional: If True, returns only labelled images.
fields : List[str], optional: List of fields to return.
per_page : int, optional: Number of images to return per page.
semaphore : asyncio.Semaphore, optional: Semaphore for limiting the number of simultaneous requests.
kwargs: Additional arguments.

Returns

List of images in dataset.

Return type

AsyncGenerator[List[ImageInfo]]

Usage example

import supervisely as sly
import asyncio

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

loop = sly.utils.get_or_create_event_loop()
images = loop.run_until_complete(api.image.get_list_async(123456, per_page=600))

async get_list_idx_page_async(method, data)¶

Get the list of items for a given page number. Page number is specified in the data dictionary.

Parameters

method : str: Method to call for listing items.
data : dict: Data to pass to the API method.

Returns

List of items.

Return type

Tuple[int, List[NamedTuple]]

async get_list_page_generator_async(method, data, pages_count=None, semaphore=None)¶

Yields list of images in dataset asynchronously page by page.

Parameters

method : str: Method to call for listing items.
data : dict: Data to pass to the API method.
pages_count : int, optional: Preferred number of pages to retrieve if used with a per_page limit. Will be automatically adjusted if the pagesCount differs from the requested number.
semaphore : asyncio.Semaphore, optional: Semaphore for limiting the number of simultaneous requests.
kwargs: Additional arguments.

Returns

List of images in dataset.

Return type

AsyncGenerator[List[ImageInfo]]

Usage example

import supervisely as sly
import asyncio

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

method = 'images.list'
data = {
    'datasetId': 123456
}

loop = sly.utils.get_or_create_event_loop()
images = loop.run_until_complete(api.image.get_list_generator_async(method, data))

get_project_id(image_id)[source]¶

Gets Project ID by Image ID.

Parameters

image_id : int: Image ID in Supervisely.

Returns

Project ID where Image is located.

Return type

int

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

img_id = 121236920
img_project_id = api.image.get_project_id(img_id)
print(img_project_id)
# Output: 53939

group_images_for_multiview(image_ids, group_name, multiview_tag_name=None)[source]¶

Group images for multi-view by tag with given name. If tag does not exist in project, will create it first.

Note:

All images must belong to the same project.
Tag must be of type ANY_STRING and applicable to images.
Recommended number of images in group is 6-12.

Parameters

image_ids : List[int]: List of Images IDs in Supervisely.
group_name : str: Group name. Images will be assigned by group tag with this value.
multiview_tag_name : str, optional: Multiview tag name in Supervisely. If None, will use default ‘multiview’ tag name. If tag does not exist in project, will create it first.

Returns

None

Return type

NoneType

Raises

ValueError – if tag is not of type ANY_STRING or not applicable to images

Usage example

# ? option 1
import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

BATCH_SIZE = 6
image_ids = [2389126, 2389127, 2389128, 2389129, 2389130, 2389131, ...]

# group images for multiview
for group_name, ids in enumerate(sly.batched(image_ids, batch_size=BATCH_SIZE)):
    api.image.group_images_for_multiview(ids, group_name)


# ? option 2 (with sly.ApiContext)
import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

BATCH_SIZE = 6
image_ids = [2389126, 2389127, 2389128, 2389129, 2389130, 2389131, ...]
project_id = 111111 # change to your project id


# * make sure that `with_settings=True` is set to get project settings from server
project_meta_json = api.project.get_meta(project_id, with_settings=True)
project_meta = sly.ProjectMeta.from_json(project_meta_json)

# create custom tag meta (optional)
multiview_tag_name = 'cars'
tag_meta = sly.TagMeta(multiview_tag_name, sly.TagValueType.ANY_STRING)
project_meta = project_meta.add_tag_meta(tag_meta)
project_meta = api.project.update_meta(project_id, project_meta) # update meta on server

# group images for multiview
with sly.ApiContext(api, project_id=project_id, project_meta=project_meta):
    for group_name, ids in enumerate(sly.batched(image_ids, batch_size=BATCH_SIZE)):
        api.image.group_images_for_multiview(ids, group_name, multiview_tag_name)

static info_sequence()[source]¶

Get list of all ImageInfo field names.

Returns: List of ImageInfo field names.`
Return type: List[str]

static info_tuple_name()[source]¶

Get string name of ImageInfo NamedTuple.

Returns: NamedTuple name.
Return type: str

move(dst_dataset_id, id, change_name_if_conflict=False, with_annotations=False)[source]¶

Moves Image with given ID to destination Dataset.

Parameters

dst_dataset_id : int: Destination Dataset ID in Supervisely.
id : int: Image ID in Supervisely.
change_name_if_conflict : bool, optional: If True adds suffix to the end of Image name when Dataset already contains an Image with identical name, If False and images with the identical names already exist in Dataset raises error.
with_annotations : bool, optional: If True Image will be copied to Dataset with annotations, otherwise only Images without annotations.

Returns

Information about Image. See info_sequence

Return type

ImageInfo

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

dst_ds_id = 365484
img_id = 533336920

img_info = api.image.copy(dst_ds_id, img_id, with_annotations=True)

move_batch(dst_dataset_id, ids, change_name_if_conflict=False, with_annotations=False, progress_cb=None)[source]¶

Moves Images with given IDs to Dataset.

Parameters

dst_dataset_id : int: Destination Dataset ID in Supervisely.
ids : List[int]: Images IDs in Supervisely.
change_name_if_conflict : bool, optional: If True adds suffix to the end of Image name when Dataset already contains an Image with identical name, If False and images with the identical names already exist in Dataset raises error.
with_annotations : bool, optional: If True Image will be copied to Dataset with annotations, otherwise only Images without annotations.
progress_cb : tqdm or callable, optional: Function for tracking the progress of moving.

Raises

TypeError if type of ids is not list

Raises

ValueError if images ids are from the destination Dataset

Returns

List with information about Images. See info_sequence

Return type

List[ImageInfo]

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

ds_lemon_id = 1780
ds_kiwi_id = 1233

ds_lemon_img_infos = api.image.get_list(ds_lemon_id)
ds_kiwi_img_infos = api.image.get_list(ds_kiwi_id)

fruit_img_ids = []
for lemon_img_info, kiwi_img_info in zip(ds_lemon_img_infos, ds_kiwi_img_infos):
    fruit_img_ids.append(lemon_img_info.id)
    fruit_img_ids.append(kiwi_img_info.id)

ds_fruit_id = 2574
ds_fruit_img_infos = api.image.move_batch(ds_fruit_id, fruit_img_ids, with_annotations=True)

move_batch_optimized(src_dataset_id, src_image_infos, dst_dataset_id, with_annotations=True, progress_cb=None, dst_names=None, batch_size=500, skip_validation=False, save_source_date=True)[source]¶

Moves Images with given IDs to Dataset.

Parameters

src_dataset_id : int: Source Dataset ID in Supervisely.
src_image_infos : List [ ImageInfo ]: ImageInfo objects of images to move.
dst_dataset_id : int: Destination Dataset ID in Supervisely.
with_annotations : bool, optional: If True Image will be copied to Dataset with annotations, otherwise only Images without annotations.
progress_cb : tqdm or callable, optional: Function for tracking the progress of moving.
dst_names : List [ ImageInfo ], optional: ImageInfo list with existing items in destination dataset.
batch_size : int, optional: Number of elements to copy for each request.
skip_validation : bool, optional: Flag for skipping additinal validations.
save_source_date : bool, optional: Save source annotation dates (creation and modification) or create a new date.

Raises

TypeError if type of src_image_infos is not list

Returns

List with information about Images. See info_sequence

Return type

List[ImageInfo]

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

src_ds_id = 2231
img_infos = api.image.get_list(src_ds_id)

dest_ds_id = 2574
dest_img_infos = api.image.move_batch_optimized(src_ds_id, img_infos, dest_ds_id)

preview_url(url, width=None, height=None, quality=70, ext='jpeg', method='auto')[source]¶

Previews Image with the given resolution parameters. Learn more about resize parameters here.

Parameters

url : str: Full Image storage URL.
width : int: Preview Image width.
height : int: Preview Image height.
quality : int: Preview Image quality.
ext : str, optional: Preview Image extension, available values: “jpeg”, “png”.
method : str, optional: Preview Image resize method, available values: “fit”, “fill”, “fill-down”, “force”, “auto”.

Returns

New URL with resized Image

Return type

str

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

image_id = 376729
img_info = api.image.get_info_by_id(image_id)
img_preview_url = api.image.preview_url(img_info.full_storage_url, width=512, height=256)

raise_name_intersections_if_exist(dataset_id, names, message=None)[source]¶

Raises error if images with given names already exist in dataset. Default error message: “Images with the following names already exist in dataset [ID={dataset_id}]: {name_intersections}. Please, rename images and try again or set change_name_if_conflict=True to rename automatically on upload.”

Parameters

dataset_id : int: Dataset ID in Supervisely.
names : List[str]: List of names to check.
message : str, optional: Error message.

Returns

None

Return type

None

remove(image_id)[source]¶

Remove image from supervisely by id. All image IDs must belong to the same dataset. Therefore, it is necessary to sort IDs before calling this method.

Parameters

image_id : int: Images ID in Supervisely.

Returns

None

Return type

NoneType

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

image_id = 2389126
api.image.remove(image_id)

remove_batch(ids, progress_cb=None, batch_size=50)[source]¶

Remove images from supervisely by IDs. IDs must belong to the same project.

Parameters

ids : List[int]: List of Images IDs in Supervisely.
progress_cb : tqdm or callable, optional: Function for tracking progress of removing.

Returns

None

Return type

NoneType

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

image_ids = [2389126, 2389127]
api.image.remove_batch(image_ids)

rename(id, name)[source]¶

Renames Image with given ID.

Parameters

id : int: Image ID in Supervisely.
name : str: New Image name.

Returns

Information about updated Image.

Return type

ImageInfo

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

image_id = 376729
new_image_name = 'new_image_name.jpg'
img_info = api.image.rename(image_id, new_image_name)

set_custom_sort(id, sort_value)[source]¶

Sets custom sort value for image with given ID.

Parameters

id : int: Image ID in Supervisely.
sort_value : str: Sort value.

Returns

json-encoded content of a response.

Return type

Dict[str, Any]

set_custom_sort_bulk(ids, sort_values)[source]¶

Sets custom sort values for images with given IDs.

Parameters

ids : List[int]: Image IDs in Supervisely.
sort_values : List[str]: List of custom sort values that will be set for images. It is stored as a key customSort value in the image meta.

Returns

json-encoded content of a response.

Return type

Dict[str, Any]

set_remote(images, links)[source]¶

This method helps to change local source to remote for images without re-uploading them as new.

Parameters

images : List[int]: List of image ids.
links : List[str]: List of remote links.

Returns

json-encoded content of a response.

Usage example

import supervisely as sly

api = sly.Api.from_env()

images = [123, 124, 125]
links = [
    "s3://bucket/lemons/ds1/img/IMG_444.jpeg",
    "s3://bucket/lemons/ds1/img/IMG_445.jpeg",
    "s3://bucket/lemons/ds1/img/IMG_446.jpeg",
]
result = api.image.set_remote(images, links)

storage_url(path_original)[source]¶

Get full Image URL link in Supervisely server.

Parameters

path_original : str: Original Image path in Supervisely server.

Returns

Full Image URL link in Supervisely server

Return type

str

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

image_id = 376729
img_info = api.image.get_info_by_id(image_id)
img_storage_url = api.image.storage_url(img_info.path_original)

static update_custom_sort(meta, custom_sort)[source]¶

Updates a copy of the meta dictionary with a new custom sort value.

Parameters

meta : Dict[str, Any]: Image meta dictionary.
custom_sort : str: Custom sort value.

Returns

Updated meta dictionary.

Return type

Dict[str, Any]

update_meta(id, meta)[source]¶

It is possible to add custom JSON data to every image for storing some additional information. Updates Image metadata by ID. Metadata is visible in Labeling Tool. Supervisely also have 2 apps: import metadata and export metadata

Parameters

id : int: Image ID in Supervisely.
meta : dict: Custom additional image info that contain image technical and/or user-generated data.

Raises

TypeError if meta type is not dict

Returns

Image information in dict format with new meta

Return type

dict

Usage example

import os
import json
import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

image_info = api.image.get_info_by_id(id=3212008)
print(image_info.meta)
# Output: {}

new_meta = {'Camera Make': 'Canon', 'Color Space': 'sRGB', 'Focal Length': '16 mm'}
new_image_info = api.image.update_meta(id=3212008, meta=new_meta)

image_info = api.image.get_info_by_id(id=3212008)
print(json.dumps(obj=image_info.meta, indent=4))
# Output: {
#     "Camera Make": "Canon",
#     "Color Space": "sRGB",
#     "Focal Length": "16 mm"
# }

update_tag_value(tag_id, value)[source]¶

Update tag value with given ID.

Parameters

tag_id : int: Tag ID in Supervisely.
value : str or float: Tag value.
project_meta : ProjectMeta: Project Meta.

Returns

Information about updated tag.

Return type

dict

Usage example

import supervisely as sly

os.environ[‘SERVER_ADDRESS’] = ‘https://app.supervisely.com’ os.environ[‘API_TOKEN’] = ‘Your Supervisely API Token’ api = sly.Api.from_env()

tag_id = 277083 new_value = ‘new_value’ api.image.update_tag_value(tag_id, new_value)

upload_blob_images(dataset, blob_file, metas=None, change_name_if_conflict=True, progress_cb=None, return_image_infos_generator=False)[source]¶

Uploads images from blob file in Team Files to dataset.

IMPORTANT: File with image offsets should be in the same directory as the blob file. This file should be named as the blob file but with the suffix _offsets.pkl. It must be a Pickle file with the BlobImageInfos that define the range of bytes representing the image in the binary. To prepare the offsets file, use the supervisely.fs.save_blob_offsets_pkl function.

Parameters

dataset : Union[DatasetInfo, int]: Dataset in Supervisely. Can be DatasetInfo object or dataset ID. It is recommended to use DatasetInfo object to avoid additional API requests.
blob_file : Union[FileInfo, str]: Blob file in Team Files. Can be FileInfo object or path to blob file. It is recommended to use FileInfo object to avoid additional API requests.
metas : Optional[List[Dict[str, Any]], optional: List of metas for images.
change_name_if_conflict : bool, optional: If True adds suffix to the end of Image name when Dataset already contains an Image with identical name, If False and images with the identical names already exist in Dataset skips them.
progress_cb : Optional[Union[tqdm, Callable]]: Function for tracking upload progress. Tracks the count of processed items.
return_image_infos_generator : bool, optional: If True, returns generator of ImageInfo objects. Otherwise, returns None.

Returns

Generator of ImageInfo objects if return_image_infos_generator is True, otherwise None.

Return type

Union[Generator[ImageInfo, None], None]

upload_by_offsets(dataset, team_file_id, names=None, offsets=None, progress_cb=None, metas=None, batch_size=50, skip_validation=False, conflict_resolution=None, validate_meta=False, use_strict_validation=False, use_caching_for_validation=False)[source]¶

Upload images from blob file in Team Files by offsets to Dataset with prepared names. To upload large number of images, use api.image.upload_by_offsets_generator() instead.

If you include metas during the upload, you can add a custom sort parameter for images. To achieve this, use the context manager api.image.add_custom_sort() with the desired key name from the meta dictionary to be used for sorting.

Parameters

dataset : Union[DatasetInfo,int]

Dataset ID or DatasetInfo object in Supervisely.

team_file_id : int

ID of the binary file in the team storage.

names : List[str], optional

Images names with extension.

REQUIRED if there is no file containing offsets in the team storage at the same level as the TAR file. Offset file must be named as the TAR file with the _offsets.pkl suffix and must be represented in pickle format. Example: tar_name_offsets.pkl

offsets : List[dict], optional

List of dictionaries with file offsets that define the range of bytes representing the image in the binary. Example: [{"offsetStart": 0, "offsetEnd": 100}, {"offsetStart": 101, "offsetEnd": 200}].

REQUIRED if there is no file containing offsets in the team storage at the same level as the TAR file. Offset file must be named as the TAR file with the _offsets.pkl suffix and must be represented in pickle format. Example: tar_name_offsets.pkl

progress_cb : tqdm or callable, optional

Function for tracking the progress of uploading.

metas : List[dict], optional

Custom additional image infos that contain images technical and/or user-generated data as list of separate dicts.

batch_size : int, optional

Number of images to upload in one batch.

skip_validation : bool, optional

Skips validation for images, can result in invalid images being uploaded.

conflict_resolution : Optional[Literal["rename", "skip", "replace"]]

The strategy to resolve upload conflicts. ‘Replace’ option will replace the existing images in the dataset with the new images. The images that are being deleted are logged. ‘Skip’ option will ignore the upload of new images that would result in a conflict. An original image’s ImageInfo list will be returned instead. ‘Rename’ option will rename the new images to prevent any conflict.

validate_meta : bool, optional

If True, validates provided meta with saved JSON schema.

use_strict_validation : bool, optional

If True, uses strict validation.

use_caching_for_validation : bool, optional

If True, uses caching for validation.

Returns

List with information about Images. See info_sequence

Return type

List[ImageInfo]

Usage example

import supervisely as sly
from supervisely.api.module_api import ApiField


server_address = 'https://app.supervisely.com'
api_token = 'Your Supervisely API Token'
api = sly.Api(server_address, api_token)

dataset_id = 452984
names = ['lemon_1.jpg', 'lemon_1.jpg']
offsets = [
    {ApiField.OFFSET_START: 0, ApiField.OFFSET_END: 100},
    {ApiField.OFFSET_START: 101, ApiField.OFFSET_END: 200}
]
team_file_id = 123456
new_imgs_info = api.image.upload_by_offsets(dataset_id, team_file_id, names, offsets,  metas)

# Output example:
#   ImageInfo(id=136281,
#             name='lemon_1.jpg',
#             link=None,
#             hash=None,
#             mime=None,
#             ext=None,
#             size=100,
#             width=None,
#             height=None,
#             labels_count=0,
#             dataset_id=452984,
#             created_at='2025-03-21T18:30:08.551Z',
#             updated_at='2025-03-21T18:30:08.551Z',
#             meta={},
#             path_original='/h5un6l2.../eyJ0eXBlIjoic291cmNlX2Jsb2I...',
#             full_storage_url='http://storage:port/h5un6l2...,
#             tags=[],
#             created_by_id=user),
#   ImageInfo(...)

upload_by_offsets_generator(dataset, team_file_id, offsets_file_path=None, progress_cb=None, metas=None, batch_size=10000, skip_validation=False, conflict_resolution=None, validate_meta=False, use_strict_validation=False, use_caching_for_validation=False)[source]¶

Upload images from blob file in Team Files by offsets to Dataset. Generates information about uploaded images in batches of max size 10000. File names will be taken from the offset file.

This method is better suited for large datasets, as it does not require resulting all the images into memory at once.

If you include metas during the upload, you can add a custom sort parameter for images. To achieve this, use the context manager api.image.add_custom_sort() with the desired key name from the meta dictionary to be used for sorting.

Parameters

dataset : Union[DatasetInfo,int]: Dataset ID or DatasetInfo object in Supervisely.
team_file_id : int: ID of the binary file in the team storage.
offsets_file_path : str, optional: Local path to the file with blob images offsets.
progress_cb : tqdm or callable, optional: Function for tracking the progress of uploading.
metas : Dict, optional: Custom additional image infos as dict where: keys - image names, values - image technical and/or user-generated data dicts
batch_size : int, optional: Number of images to upload in one batch.
skip_validation : bool, optional: Skips validation for images, can result in invalid images being uploaded.
conflict_resolution : Optional[Literal["rename", "skip", "replace"]]: The strategy to resolve upload conflicts. ‘Replace’ option will replace the existing images in the dataset with the new images. The images that are being deleted are logged. ‘Skip’ option will ignore the upload of new images that would result in a conflict. An original image’s ImageInfo list will be returned instead. ‘Rename’ option will rename the new images to prevent any conflict.
validate_meta : bool, optional: If True, validates provided meta with saved JSON schema.
use_strict_validation : bool, optional: If True, uses strict validation.
use_caching_for_validation : bool, optional: If True, uses caching for validation.

Returns

Generator with information about Images. See ImageInfo

Return type

Generator[ImageInfo, None, None]

Usage example

import supervisely as sly
from supervisely.api.module_api import ApiField


server_address = 'https://app.supervisely.com'
api_token = 'Your Supervisely API Token'
api = sly.Api(server_address, api_token)

dataset_id = 452984
team_file_id = 123456
img_infos = []
new_imgs_info_generator = api.image.upload_by_offsets_generator(dataset_id, team_file_id)
for img_infos_batch in new_imgs_info_generator:
    img_infos.extend(img_infos_batch)

upload_dir(dataset_id, dir_path, recursive=True, change_name_if_conflict=True, progress_cb=None)[source]¶

Uploads all images with supported extensions from given directory to Supervisely. Optionally, uploads images from subdirectories of given directory.

Parameters

dataset_id : int: Dataset ID in Supervisely.
dir_path : str: Path to directory with images.
recursive : bool, optional: If True uploads images from subdirectories of given directory recursively, otherwise only images from given directory.
change_name_if_conflict : bool, optional: If True adds suffix to the end of Image name when Dataset already contains an Image with identical name, If False and images with the identical names already exist in Dataset raises error.
progress_cb : Optional[Union[tqdm, Callable]]: Function for tracking upload progress.

Returns

List of uploaded images infos

Return type

List[ImageInfo]

upload_dirs(dataset_id, dir_paths, recursive=True, change_name_if_conflict=True, progress_cb=None)[source]¶

Uploads all images with supported extensions from given directories to Supervisely. Optionally, uploads images from subdirectories of given directories.

Parameters

dataset_id : int: Dataset ID in Supervisely.
dir_paths : List[str]: List of paths to directories with images.
recursive : bool, optional: If True uploads images from subdirectories of given directories recursively, otherwise only images from given directories.
change_name_if_conflict : bool, optional: If True adds suffix to the end of Image name when Dataset already contains an Image with identical name, If False and images with the identical names already exist in Dataset raises error.
progress_cb : Optional[Union[tqdm, Callable]]: Function for tracking upload progress.

Returns

List of uploaded images infos

Return type

List[ImageInfo]

upload_hash(dataset_id, name, hash, meta=None)[source]¶

Upload Image from given hash to Dataset.

If you include meta during the upload, you can add a custom sort parameter for image. To achieve this, use the context manager api.image.add_custom_sort() with the desired key name from the meta dictionary to be used for sorting. Refer to the example section for more details.

Parameters

dataset_id : int: Dataset ID in Supervisely.
name : str: Image name with extension.
hash : str: Image hash.
meta : dict, optional: Custom additional image info that contain image technical and/or user-generated data.

Returns

Information about Image. See info_sequence

Return type

ImageInfo

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

dst_dataset_id = 452984
im_info = api.image.get_info_by_id(193940090)
hash = im_info.hash
# It is necessary to upload image with the same name(extention) as in src dataset
name = im_info.name
meta = {1: 'meta_example'}
new_in_info = api.image.upload_hash(dst_dataset_id, name, hash, meta)
print(json.dumps(new_in_info, indent=4))
# Output: [
#     196793586,
#     "IMG_0748.jpeg",
#     null,
#     "NEjmnmdd7DOzaFAKK/nCIl5CtcwZeMkhW3CHe875p9g=",
#     "image/jpeg",
#     "jpeg",
#     66885,
#     600,
#     500,
#     0,
#     452984,
#     "2021-03-16T09:09:45.587Z",
#     "2021-03-16T09:09:45.587Z",
#     {
#         "1": "meta_example"
#     },
#     "/h5un6l2bnaz1vj8a9qgms4-public/images/original/P/a/kn/W2mzMQg435d6wG0.jpg",
#     "https://app.supervise.ly/h5un6l2bnaz1vj8a9qgms4-public/images/original/P/a/kn/W2mzMQg435hiHJAPgMU.jpg"
# ]

# Add custom sort parameter for image
new_dataset_id = 452985
im_info = api.image.get_info_by_id(193940090)
print(im_info.meta)
# Output: {'my-key':'a'}
with api.image.add_custom_sort(key="my-key"):
    img_info = api.image.upload_hash(new_dataset_id, name=im_info.name, hash=im_info.hash, meta=im_info.meta)

upload_hashes(dataset_id, names, hashes, progress_cb=None, metas=None, batch_size=50, skip_validation=False, conflict_resolution=None, validate_meta=False, use_strict_validation=False, use_caching_for_validation=False)[source]¶

Upload images from given hashes to Dataset.

If you include metas during the upload, you can add a custom sort parameter for images. To achieve this, use the context manager api.image.add_custom_sort() with the desired key name from the meta dictionary to be used for sorting. Refer to the example section for more details.

Parameters

dataset_id : int: Dataset ID in Supervisely.
names : List[str]: Images names with extension.
hashes : List[str]: Images hashes.
progress_cb : tqdm or callable, optional: Function for tracking the progress of uploading.
metas : List[dict], optional: Custom additional image infos that contain images technical and/or user-generated data as list of separate dicts.
batch_size : int, optional: Number of images to upload in one batch.
skip_validation : bool, optional: Skips validation for images, can result in invalid images being uploaded.
conflict_resolution : Optional[Literal["rename", "skip", "replace"]]: The strategy to resolve upload conflicts. ‘Replace’ option will replace the existing images in the dataset with the new images. The images that are being deleted are logged. ‘Skip’ option will ignore the upload of new images that would result in a conflict. An original image’s ImageInfo list will be returned instead. ‘Rename’ option will rename the new images to prevent any conflict.
validate_meta : bool, optional: If True, validates provided meta with saved JSON schema.
use_strict_validation : bool, optional: If True, uses strict validation.
use_caching_for_validation : bool, optional: If True, uses caching for validation.

Returns

List with information about Images. See info_sequence

Return type

List[ImageInfo]

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

src_dataset_id = 447130
hashes = []
names = []
metas = []
imgs_info = api.image.get_list(src_dataset_id)
# Create lists of hashes, images names and meta information for each image
for im_info in imgs_info:
    hashes.append(im_info.hash)
    # It is necessary to upload images with the same names(extentions) as in src dataset
    names.append(im_info.name)
    metas.append({im_info.name: im_info.size})

dst_dataset_id = 452984
progress = sly.Progress("Images upload: ", len(hashes))
new_imgs_info = api.image.upload_hashes(dst_dataset_id, names, hashes, progress.iters_done_report, metas)
# Output:
# {"message": "progress", "event_type": "EventType.PROGRESS", "subtask": "Images downloaded: ", "current": 0, "total": 10, "timestamp": "2021-03-16T11:59:07.444Z", "level": "info"}
# {"message": "progress", "event_type": "EventType.PROGRESS", "subtask": "Images downloaded: ", "current": 10, "total": 10, "timestamp": "2021-03-16T11:59:07.644Z", "level": "info"}

# Add custom sort parameter for images
new_dataset_id = 452985
new_metas = [{'my-key':'a'}, {'my-key':'b'}, {'my-key':'c'}]
with api.image.add_custom_sort(key="my-key"):
    img_infos = api.image.upload_hashes(new_dataset_id, names=names, hashes=hashes, metas=new_metas)

upload_id(dataset_id, name, id, meta=None)[source]¶

Upload Image by ID to Dataset.

If you include meta during the upload, you can add a custom sort parameter for image. To achieve this, use the context manager api.image.add_custom_sort() with the desired key name from the meta dictionary to be used for sorting. Refer to the example section for more details.

Parameters

dataset_id : int: Destination Dataset ID in Supervisely.
name : str: Image name with extension.
id : int: Source image ID in Supervisely.
meta : dict, optional: Custom additional image info that contain image technical and/or user-generated data.

Returns

Information about Image. See info_sequence

Return type

ImageInfo

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

dst_dataset_id = 452984
im_info = api.image.get_info_by_id(193940090)
id = im_info.id
# It is necessary to upload image with the same name(extention) as in src dataset
name = im_info.name
meta = {1: 'meta_example'}
new_in_info = api.image.upload_id(dst_dataset_id, name, id, meta)
print(json.dumps(new_in_info, indent=4))
# Output: [
#     196793605,
#     "IMG_0748.jpeg",
#     null,
#     "NEjmnmdd7DOzaFAKK/nCIl5CtcwZeMkhW3CHe875p9g=",
#     "image/jpeg",
#     "jpeg",
#     66885,
#     600,
#     500,
#     0,
#     452984,
#     "2021-03-16T09:27:12.620Z",
#     "2021-03-16T09:27:12.620Z",
#     {
#         "1": "meta_example"
#     },
#     "/h5un6l2bnaz1vj8a9qgms4-public/images/original/P/a/kn/W2mzMQg435d6wG0AJGJTOsL1FqMUNOPqu4VdzFAN36LqtGwBIE4AmLOQ1BAxuIyB0bHJAPgMU.jpg",
#     "https://app.supervise.ly/h5un6l2bnaz1vj8a9qgms4-public/images/original/P/a/kn/iEaDEkejnfnb1Tz56ka0hiHJAPgMU.jpg"
# ]

# Add custom sort parameter for image
new_dataset_id = 452985
im_info = api.image.get_info_by_id(193940090)
print(im_info.meta)
# Output: {"my-key": "a"}
with api.image.add_custom_sort(key="my-key"):
    img_info = api.image.upload_id(new_dataset_id, name=im_info.name, id=im_info.id, meta=im_info.meta)

upload_ids(dataset_id, names, ids, progress_cb=None, metas=None, batch_size=50, force_metadata_for_links=True, infos=None, skip_validation=False, conflict_resolution=None)[source]¶

Upload Images by IDs to Dataset.

If you include metas during the upload, you can add a custom sort parameter for images. To achieve this, use the context manager api.image.add_custom_sort() with the desired key name from the meta dictionary to be used for sorting. Refer to the example section for more details.

Parameters

dataset_id : int: Destination Dataset ID in Supervisely.
names : List[str]: Source images names with extension.
ids : List[int]: Images IDs.
progress_cb : tqdm or callable, optional: Function for tracking the progress of uploading.
metas : List[dict], optional: Custom additional image infos that contain images technical and/or user-generated data as list of separate dicts.
batch_size : int, optional: Number of images to upload in one batch.
force_metadata_for_links : bool, optional: Calculate metadata for links. If False, metadata will be empty.
infos : List[ImageInfo], optional: DEPRECATED: This parameter is not used.
skip_validation : bool, optional: Skips validation for images, can result in invalid images being uploaded.
conflict_resolution : Optional[Literal["rename", "skip", "replace"]]: The strategy to resolve upload conflicts. ‘Replace’ option will replace the existing images in the dataset with the new images. The images that are being deleted are logged. ‘Skip’ option will ignore the upload of new images that would result in a conflict. An original image’s ImageInfo list will be returned instead. ‘Rename’ option will rename the new images to prevent any conflict.

Returns

List with information about Images. See info_sequence

Return type

List[ImageInfo]

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

src_dataset_id = 447130

ids = []
names = []
metas = []
imgs_info = api.image.get_list(src_dataset_id)
# Create lists of ids, images names and meta information for each image
for im_info in imgs_info:
    ids.append(im_info.id)
    # It is necessary to upload images with the same names(extentions) as in src dataset
    names.append(im_info.name)
    metas.append({im_info.name: im_info.size})

dst_dataset_id = 452984
progress = sly.Progress("Images upload: ", len(ids))
new_imgs_info = api.image.upload_ids(dst_dataset_id, names, ids, progress.iters_done_report, metas)
# Output:
# {"message": "progress", "event_type": "EventType.PROGRESS", "subtask": "Images downloaded: ", "current": 0, "total": 10, "timestamp": "2021-03-16T12:31:36.550Z", "level": "info"}
# {"message": "progress", "event_type": "EventType.PROGRESS", "subtask": "Images downloaded: ", "current": 10, "total": 10, "timestamp": "2021-03-16T12:31:37.119Z", "level": "info"}

# Add custom sort parameter for images
new_dataset_id = 452985
new_metas = [{'my-key':'a'}, {'my-key':'b'}, {'my-key':'c'}]
with api.image.add_custom_sort(key="my-key"):
    img_infos = api.image.upload_ids(new_dataset_id, names=names, ids=ids, metas=new_metas)

upload_link(dataset_id, name, link, meta=None, force_metadata_for_links=True)[source]¶

Uploads Image from given link to Dataset.

If you include meta during the upload, you can add a custom sort parameter for image. To achieve this, use the context manager api.image.add_custom_sort() with the desired key name from the meta dictionary to be used for sorting. Refer to the example section for more details.

Parameters

dataset_id : int: Dataset ID in Supervisely.
name : str: Image name with extension.
link : str: Link to Image.
meta : dict, optional: Custom additional image info that contain image technical and/or user-generated data.
force_metadata_for_links : bool, optional: Calculate metadata for link. If False, metadata will be empty.

Returns

Information about Image. See info_sequence

Return type

ImageInfo

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

img_name = 'Avatar.jpg'
img_link = 'https://m.media-amazon.com/images/M/MV5BMTYwOTEwNjAzMl5BMl5BanBnXkFtZTcwODc5MTUwMw@@._V1_.jpg'

img_info = api.image.upload_link(dataset_id, img_name, img_link)

# Add custom sort parameter for image
img_meta = {"my-key": "a"}
with api.image.add_custom_sort(key="my-key"):
    img_info = api.image.upload_link(dataset_id, img_name, img_link, meta=img_meta)

upload_links(dataset_id, names, links, progress_cb=None, metas=None, batch_size=50, force_metadata_for_links=True, skip_validation=False, conflict_resolution=None)[source]¶

Uploads Images from given links to Dataset.

If you include metas during the upload, you can add a custom sort parameter for images. To achieve this, use the context manager api.image.add_custom_sort() with the desired key name from the meta dictionary to be used for sorting. Refer to the example section for more details.

Parameters

dataset_id : int: Dataset ID in Supervisely.
names : List[str]: Images names with extension.
links : List[str]: Links to Images.
progress_cb : tqdm or callable, optional: Function for tracking the progress of uploading.
metas : List[dict], optional: Custom additional image infos that contain images technical and/or user-generated data as list of separate dicts.
force_metadata_for_links : bool, optional: Calculate metadata for links. If False, metadata will be empty.
skip_validation : bool, optional: Skips validation for images, can result in invalid images being uploaded.
conflict_resolution : Optional[Literal["rename", "skip", "replace"]]: The strategy to resolve upload conflicts. ‘Replace’ option will replace the existing images in the dataset with the new images. The images that are being deleted are logged. ‘Skip’ option will ignore the upload of new images that would result in a conflict. An original image’s ImageInfo list will be returned instead. ‘Rename’ option will rename the new images to prevent any conflict.

Returns

List with information about Images. See info_sequence

Return type

List[ImageInfo]

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

img_names = ['Avatar.jpg', 'Harry Potter.jpg', 'Avengers.jpg']
img_links = ['https://m.media-amazon.com/images/M/MV5BMTYwOTEwNjAzMl5BMl5BanBnXkFtZTcwODc5MTUwMw@@._V1_.jpg',
             'https://m.media-amazon.com/images/M/MV5BNDYxNjQyMjAtNTdiOS00NGYwLWFmNTAtNThmYjU5ZGI2YTI1XkEyXkFqcGdeQXVyMTMxODk2OTU@._V1_.jpg',
             'https://m.media-amazon.com/images/M/MV5BNjQ3NWNlNmQtMTE5ZS00MDdmLTlkZjUtZTBlM2UxMGFiMTU3XkEyXkFqcGdeQXVyNjUwNzk3NDc@._V1_.jpg']

img_infos = api.image.upload_links(dataset_id, img_names, img_links)

# Add custom sort parameter for images
img_metas = [{'my-key':'a'}, {'my-key':'b'}, {'my-key':'c'}]
with api.image.add_custom_sort(key="my-key"):
    img_infos = api.image.upload_links(dataset_id, names=img_names, links=img_links, metas=img_metas)

upload_medical_images(dataset_id, paths, group_tag_name=None, metas=None, progress_cb=None)[source]¶

Upload medical 2D images (DICOM) to Supervisely and group them by specified or default tag.

If you include metas during the upload, you can add a custom sort parameter for images. To achieve this, use the context manager api.image.add_custom_sort() with the desired key name from the meta dictionary to be used for sorting. Refer to the example section for more details.

Parameters

dataset_id : int: Dataset ID in Supervisely.
paths : List[str]: List of paths to images.
group_tag_name : str, optional: Group name. All images will be assigned by tag with this group name. If group_tag_name is None, the images will be grouped by one of the default tags.
metas : List[Dict], optional: Custom additional image infos that contain images technical and/or user-generated data as list of separate dicts.
progress_cb : tqdm or callable, optional: Function for tracking upload progress.

Returns

List of uploaded images infos.

Return type

List[ImageInfo]

Raises

Exception – If tag does not exist in project or tag is not of type ANY_STRING
Exception – If length of metas is not equal to the length of paths.

Usage example

import os
from dotenv import load_dotenv
from tqdm import tqdm

import supervisely as sly

# Load secrets and create API object from .env file (recommended)
# Learn more here: https://developer.supervisely.com/getting-started/basics-of-authentication
if sly.is_development():
   load_dotenv(os.path.expanduser("~/supervisely.env"))

api = sly.Api.from_env()

dataset_id = 123456
paths = ['path/to/medical_01.dcm', 'path/to/medical_02.dcm']
metas = [{'meta':'01'}, {'meta':'02'}]
group_tag_name = 'StudyInstanceUID'

pbar = tqdm(desc="Uploading images", total=len(paths))
image_infos = api.image.upload_medical_images(dataset_id, paths, group_tag_name, metas)

# Add custom sort parameter for images
metas = [{'my-key':'a'}, {'my-key':'b'}]
with api.image.add_custom_sort(key="my-key"):
    image_infos = api.image.upload_medical_images(dataset_id, paths, group_tag_name, metas)

upload_multispectral(dataset_id, image_name, channels=None, rgb_images=None, progress_cb=None)[source]¶

Uploads multispectral image to Supervisely, if channels are provided, they will be uploaded as separate images. If rgb_images are provided, they will be uploaded without splitting into channels as RGB images.

Parameters

dataset_id : int: dataset ID to upload images to
image_name : str: name of the image with extension.
channels : List[np.ndarray], optional: list of numpy arrays with image channels
rgb_images : List[str], optional: list of paths to RGB images which will be uploaded as is
progress_cb : tqdm or callable, optional: function for tracking upload progress

Returns

list of uploaded images infos

Return type

List[ImageInfo]

Usage example

import os
import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'

# Load secrets and create API object from .env file (recommended)
# Learn more here: https://developer.supervisely.com/getting-started/basics-of-authentication
load_dotenv(os.path.expanduser("~/supervisely.env"))

api = sly.Api.from_env()

image_name = "demo1.png"
image = cv2.imread(f"demo_data/{image_name}")

# Extract channels as 2d numpy arrays: channels = [a, b, c]
channels = [image[:, :, i] for i in range(image.shape[2])]

image_infos = api.image.upload_multispectral(api, dataset.id, image_name, channels)

upload_multiview_images(dataset_id, group_name, paths=None, metas=None, progress_cb=None, links=None, conflict_resolution='rename', force_metadata_for_links=False)[source]¶

Uploads images to Supervisely and adds a tag to them. At least one of paths or links must be provided.

If you include metas during the upload, you can add a custom sort parameter for images. To achieve this, use the context manager api.image.add_custom_sort() with the desired key name from the meta dictionary to be used for sorting. Refer to the example section for more details.

Parameters

dataset_id : int

Dataset ID in Supervisely.

tag_name : str

Tag name in Supervisely. If tag does not exist in project, create it first. Tag must be of type ANY_STRING.

group_name : str

Group name. All images will be assigned by tag with this group name.

paths : List[str]

List of paths to images.

metas : Optional[List[Dict]]

Custom additional image infos that contain images technical and/or user-generated data as list of separate dicts.

progress_cb : Optional[Union[tqdm, Callable]]

Function for tracking upload progress.

links : Optional[List[str]]

List of links to images.

conflict_resolution : Optional[Literal["rename", "skip", "replace"]]

The strategy to resolve upload conflicts. Options:

’replace’: Replaces the existing images in the dataset with the new ones if there is a conflict and logs the deletion of existing images.

’skip’: Ignores uploading the new images if there is a conflict; the original image’s ImageInfo list will be returned instead.

’rename’: (default) Renames the new images to prevent name conflicts.

force_metadata_for_links : Optional[bool]

Specifies whether to force retrieving metadata for images from links. If False, metadata fields in the response can be empty (if metadata has not been retrieved yet).

Returns

List of uploaded images infos

Return type

List[ImageInfo]

Raises

Exception – if tag does not exist in project or tag is not of type ANY_STRING

Usage example

import os
from dotenv import load_dotenv

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'

# Load secrets and create API object from .env file (recommended)
# Learn more here: https://developer.supervisely.com/getting-started/basics-of-authentication
load_dotenv(os.path.expanduser("~/supervisely.env"))

api = sly.Api.from_env()

dataset_id = 123456
paths = ['path/to/audi_01.png', 'path/to/audi_02.png']
group_name = 'audi'

image_infos = api.image.upload_multiview_images(dataset_id, group_name, paths)

# Add custom sort parameter for images
metas = [{'my-key':'a'}, {'my-key':'b'}]
with api.image.add_custom_sort(key="my-key"):
    image_infos = api.image.upload_multiview_images(dataset_id, group_name, paths, metas)

upload_np(dataset_id, name, img, meta=None)[source]¶

Upload given Image in numpy format with given name to Dataset.

If you include meta during the upload, you can add a custom sort parameter for image. To achieve this, use the context manager api.image.add_custom_sort() with the desired key name from the meta dictionary to be used for sorting. Refer to the example section for more details.

Parameters

dataset_id : int: Dataset ID in Supervisely.
name : str: Image name with extension.
img : np.ndarray: image in RGB format(numpy matrix)
meta : dict, optional: Custom additional image info that contain image technical and/or user-generated data.

Returns

Information about Image. See info_sequence

Return type

ImageInfo

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

img_np = sly.image.read("/home/admin/Downloads/7777.jpeg")
img_info = api.image.upload_np(dataset_id, name="7777.jpeg", img=img_np)

# Add custom sort parameter for image
img_meta = {'my-key':'a'}
with api.image.add_custom_sort(key="my-key"):
    img_info = api.image.upload_np(dataset_id, name="7777.jpeg", img=img_np, meta=img_meta)

upload_nps(dataset_id, names, imgs, progress_cb=None, metas=None, conflict_resolution=None)[source]¶

Upload given Images in numpy format with given names to Dataset.

If you include metas during the upload, you can add a custom sort parameter for images. To achieve this, use the context manager api.image.add_custom_sort() with the desired key name from the meta dictionary to be used for sorting. Refer to the example section for more details.

Parameters

dataset_id : int: Dataset ID in Supervisely.
names : List[str]: Images names with extension.
imgs : List[np.ndarray]: Images in RGB numpy matrix format
progress_cb : tqdm or callable, optional: Function for tracking the progress of uploading.
metas : List[dict], optional: Custom additional image infos that contain images technical and/or user-generated data as list of separate dicts.
conflict_resolution : Optional[Literal["rename", "skip", "replace"]]: The strategy to resolve upload conflicts. ‘Replace’ option will replace the existing images in the dataset with the new images. The images that are being deleted are logged. ‘Skip’ option will ignore the upload of new images that would result in a conflict. An original image’s ImageInfo list will be returned instead. ‘Rename’ option will rename the new images to prevent any conflict.

Returns

List with information about Images. See info_sequence

Return type

List[ImageInfo]

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

img_np_1 = sly.image.read("/home/admin/Downloads/7777.jpeg")
img_np_2 = sly.image.read("/home/admin/Downloads/8888.jpeg")
img_np_3 = sly.image.read("/home/admin/Downloads/9999.jpeg")

img_names = ["7777.jpeg", "8888.jpeg", "9999.jpeg"]
img_nps = [img_np_1, img_np_2, img_np_3]

img_infos = api.image.upload_nps(dataset_id, names=img_names, imgs=img_nps)

# Add custom sort parameter for images
img_metas = [{'my-key':'a'}, {'my-key':'b'}, {'my-key':'c'}]
with api.image.add_custom_sort(key="my-key"):
    img_infos = api.image.upload_nps(dataset_id, names=img_names, imgs=img_nps, metas=img_metas)

upload_path(dataset_id, name, path, meta=None, validate_meta=False, use_strict_validation=False, use_caching_for_validation=False)[source]¶

Uploads Image with given name from given local path to Dataset.

If you include meta during the upload, you can add a custom sort parameter for image. To achieve this, use the context manager api.image.add_custom_sort() with the desired key name from the meta dictionary to be used for sorting. Refer to the example section for more details.

Parameters

dataset_id : int: Dataset ID in Supervisely.
name : str: Image name with extension.
path : str: Local Image path.
meta : dict, optional: Custom additional image info that contain image technical and/or user-generated data.
validate_meta : bool, optional: If True, validates provided meta with saved JSON schema.
use_strict_validation : bool, optional: If True, uses strict validation.
use_caching_for_validation : bool, optional: If True, uses caching for validation.

Returns

Information about Image. See info_sequence

Return type

ImageInfo

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

img_info = api.image.upload_path(dataset_id, name="7777.jpeg", path="/home/admin/Downloads/7777.jpeg")

# Add custom sort parameter for image
img_meta = {'my-key':'a'}
with api.image.add_custom_sort(key="my-key"):
    img_info = api.image.upload_path(dataset_id, name="7777.jpeg", path="/home/admin/Downloads/7777.jpeg", meta=img_meta)

upload_paths(dataset_id, names, paths, progress_cb=None, metas=None, conflict_resolution=None, validate_meta=False, use_strict_validation=False, use_caching_for_validation=False)[source]¶

Uploads Images with given names from given local path to Dataset.

If you include metas during the upload, you can add a custom sort parameter for images. To achieve this, use the context manager api.image.add_custom_sort() with the desired key name from the meta dictionary to be used for sorting. Refer to the example section for more details.

Parameters

dataset_id : int: Dataset ID in Supervisely.
names : List[str]: List of Images names with extension.
paths : List[str]: List of local Images pathes.
progress_cb : tqdm or callable, optional: Function for tracking the progress of uploading.
metas : List[dict], optional: Custom additional image infos that contain images technical and/or user-generated data as list of separate dicts.
conflict_resolution : Optional[Literal["rename", "skip", "replace"]]: The strategy to resolve upload conflicts. ‘Replace’ option will replace the existing images in the dataset with the new images. The images that are being deleted are logged. ‘Skip’ option will ignore the upload of new images that would result in a conflict. An original image’s ImageInfo list will be returned instead. ‘Rename’ option will rename the new images to prevent any conflict.
validate_meta : bool, optional: If True, validates provided meta with saved JSON schema.
use_strict_validation : bool, optional: If True, uses strict validation.
use_caching_for_validation : bool, optional: If True, uses caching for validation.

Raises

ValueError if len(names) != len(paths)

Returns

List with information about Images. See info_sequence

Return type

List[ImageInfo]

Usage example

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

img_names = ["7777.jpeg", "8888.jpeg", "9999.jpeg"]
image_paths = ["/home/admin/Downloads/img/770918.jpeg", "/home/admin/Downloads/img/770919.jpeg", "/home/admin/Downloads/img/770920.jpeg"]

img_infos = api.image.upload_paths(dataset_id, names=img_names, paths=img_paths)

# Add custom sort parameter for images
img_metas = [{'my-key':'a'}, {'my-key':'b'}, {'my-key':'c'}]
with api.image.add_custom_sort(key="my-key"):
    img_infos = api.image.upload_paths(dataset_id, names=img_names, paths=img_paths, metas=img_metas)

url(team_id, workspace_id, project_id, dataset_id, image_id)[source]¶

Gets Image URL by ID.

Parameters

team_id : int: Team ID in Supervisely.
workspace_id : int: Workspace ID in Supervisely.
project_id : int: Project ID in Supervisely.
dataset_id : int: Dataset ID in Supervisely.
image_id : int: Image ID in Supervisely.

Returns

Image URL

Return type

str

Usage example

import supervisely as sly

os.environ['SERVER_ADDRESS'] = 'https://app.supervisely.com'
os.environ['API_TOKEN'] = 'Your Supervisely API Token'
api = sly.Api.from_env()

team_id = 16087
workspace_id = 23821
project_id = 53939
dataset_id = 254737
image_id = 121236920

img_url = api.image.url(team_id, workspace_id, project_id, dataset_id, image_id)
print(url)
# Output: https://app.supervise.ly/app/images/16087/23821/53939/254737#image-121236920