[Feature] Add multi file backends to imread/imwrite.

[274869388]

Motivation

The motivation of the PR is to read image from different file backends and write image to different file backends but do not need to modify the existing interface.

It includes two parts:

• Add FileClient to image io function to support different file backends.
• Add unittest to tests/test_image/test_io.py.

Modification

1. imread

   • Add a new parameter file_client_args to load which makes it support loading object from different file backends, but the parameter is optional
   • Add FileClient to read image bytes of different file backends and use imfrombytes to decode image.
   • Add tifffile backends to imfrombytes.

2. imwrite

   • Add a new parameter file_client_args to load which makes it support loading object from different file backends, but the parameter is optional
   • Add cv2.imencode to encode image and use FileClient to write image bytes to different file backends.
   • Warn when image writing fails.

3. test_io.py

   • Add MagicMock to test imread and imwrite.

BC-breaking (Optional)

No

Use cases (Optional)

import mmcv

disk_path = '/path/of/your/file'
img = mmcv.imread(disk_path)
mmcv.imwrite(img, disk_path)

s3_path = 's3://your_bucket/img.png'
img = mmcv.imread(s3_path)
img = mmcv.imread(s3_path, file_client_args={'backend': 'petrel'})
img = mmcv.imread(s3_path, file_client_args={'prefix': 's3'})
mmcv.imwrite(img, s3_path)
mmcv.imwrite(img, s3_path, file_client_args={'backend': 'petrel'})
mmcv.imwrite(img, s3_path, file_client_args={'prefix': 's3'})

http_path = 'http://path/of/your/file'
img = mmcv.imread(http_path)
img = mmcv.imread(http_path, file_client_args={'backend': 'http'})
img = mmcv.imread(http_path, file_client_args={'prefix': 'http'})
mmcv.imwrite(img, http_path)
...

Test

• disk backend works as expected
• petrel backend works as expected

Checklist

Before PR:

• I have read and followed the workflow indicated in the CONTRIBUTING.md to create this PR.
• Pre-commit or linting tools indicated in CONTRIBUTING.md are used to fix the potential lint issues.
• Bug fixes are covered by unit tests, the case that causes the bug should be added in the unit tests.
• New functionalities are covered by complete unit tests. If not, please add more unit test to ensure the correctness.
• The documentation has been modified accordingly, including docstring or example tutorials.

After PR:

• If the modification has potential influence on downstream or other related projects, this PR should be tested with some of those projects, like MMDet or MMCls.
• CLA has been signed and all committers have signed the CLA in this PR.

[CLAassistant]

All committers have signed the CLA.

[teamwong111]

LGTM but with some slight doubt.

1. Is it necessary to add some novel unittests for imwrite with other file backend?
2. Shall we refactor test file which looks like not elegant now?
3. Shall we impl exists function for all File backend before merging this PR?

[274869388]

LGTM but with some slight doubt.

1. Is it necessary to add some novel unittests for imwrite with other file backend?
2. Shall we refactor test file which looks like not elegant now?
3. Shall we impl exists function for all File backend before merging this PR?

1. I think there is no need to test other file backends because they have been fully tested in test_fileio.py.
2. I may not be ready to refactor the test file for now.
3. We plan to impl the exists function for all file backend in the future,but not now.

[teamwong111]

LGTM

[zhouzaida]

Maybe we can also add multi-file backends support for the flowread and flowwrite

[MeowZheng]

Maybe we can also add multi-file backends support for the flowread and flowwrite

Of course, but I think we should create another pr for it. I will take this work very recently.