Thank you very much for the open source code, it's a very interesting work!
You mention in your paper that your Thumbnail.jpg contains image-specific quantization tables QL and QC as well as DCT qualization coefficients C~ for the YCbCr three channels.
Is the .jpg file obtained from the original JPEG encoding the one that contains only the DCT qualization coefficients C~ and no quantization table? If so, then why does the Thumbnail.jpg in this paper additionally contain information about the quantization table but its size is still comparable to the .jpg file obtained from the original JPEG encoding?
Looking forward to your early reply.
Thank you very much for the open source code, it's a very interesting work!
You mention in your paper that your Thumbnail.jpg contains image-specific quantization tables QL and QC as well as DCT qualization coefficients C~ for the YCbCr three channels.
Is the .jpg file obtained from the original JPEG encoding the one that contains only the DCT qualization coefficients C~ and no quantization table? If so, then why does the Thumbnail.jpg in this paper additionally contain information about the quantization table but its size is still comparable to the .jpg file obtained from the original JPEG encoding?
Looking forward to your early reply.