Add byte images

[zhiltsov-max]

Summary

An optimization for images to avoid their decoding and encoding when it is unnecessary, but the data comes not from files on disk, e.g. video frames or downloaded data. In case of jpeg this allows to keep source quality.

Needed to implement raw data copying in CVAT (cvat-ai/cvat#2229)

• Added ByteImage class to represent encoded images
• Added image data checks in tests, by default enabled when data is available
• Fixed wrong indentation in VOC format converter, which led to extra disk writes for each image
• Removed image/depth field support in VOC converter to avoid unnecessary image reading
• Fixed image extension comparison when extensions in different cases

How to test

Unit tests

Checklist

• I submit my changes into the develop branch
• I have added description of my changes into CHANGELOG
• I have updated the documentation accordingly
• I have added tests to cover my changes
• I have linked related issues)

License

• I submit my code changes under the same MIT License that covers the project.
  Feel free to contact the maintainers if that's a concern.
• I have updated the license header for each file (see an example below)

# Copyright (C) 2020 Intel Corporation
#
# SPDX-License-Identifier: MIT