Implementation for technique described in this paper: reversible data hiding

On windows:

cmake -G "Visual Studio 16 2019" -S . -B build -DBENCHMARK_ENABLE_GTEST_TESTS=OFF -DBOOST_ROOT=<path to boost root> -DBOOST_LIBRARYDIR=<boost library dir>
cmake --build build --config Release

1. Example encryption: rdh.exe --mode encrypt --image-path ./lena_gray.bmp --result-path ./images/lena_gray-enc.bmp --enc-key-file ./example_encrypt_key.bin

2. Example decryption: rdh.exe --mode decrypt --image-path ./lena_gray-enc.bmp --result-path ./images/lena_gray-dec.bmp --enc-key-file ./example_encrypt_key.bin

3. Example data embedding: rdh.exe --mode embed --image-path ./lena_gray-enc.bmp --result-path ./images/lena_gray-enc-embed.bmp --embed-key-file ./example_embed_key.bin --data-file ./example_data_to_embed.bin

4. Example data extraction: ./rdh.exe --mode extract --image-path ./marked-encrypted.bmp --result-path-data ./extracted.bin --embed-key FFDDEE

5. Example image recovery: rdh.exe --mode extract --image-path ./marked-encrypted.bmp --result-path ./extracted.bmp --encryption-key AABBCC

6. Example data extraction and original image recovery: rdh.exe --mode extract --image-path ./marked-encrypted.bmp --result-path ./extracted.bmp --result-path-data ./extracted.bin --encryption-key AABBCC --embed-key FFDDEE

7. Example SSIM calculation: rdh.exe --mode ssim --image-path ./recovered_image_man.bmp --second-image ./man_original.bmp

8. Example PSNR calculation: rdh.exe --mode psnr --image-path ./recovered_image_man.bmp --second-image ./man_original.bmp

• BOOST - program_options, dynamic_bitset, sha1
• Cimg - BMP opening/saving/showing
• eigen - math operations
• gtest - google testing library
• google benchmark - benchmarking library

@article{article,
    author = {Qin, Chuan and He, Zhihong and Luo, Xiangyang and Dong, Jing},
    year = {2018},
    month = {07},
    pages = {},
    title = {Reversible Data Hiding in Encrypted Image with Separable Capability and High Embedding Capacity},
    volume = {465},
    journal = {Information Sciences},
    doi = {10.1016/j.ins.2018.07.021}
}