I am making a simple demo in Matlab. I want to extract edge maps from video sequences then encode both frame & edge maps to bit stream. Then I will decode bitstream to have frame & edge maps, and leverage edge maps to upscaling that frame.
This is not a new idea, I used huffmanenco to encode image, but I'm facing some difficulties to encode edge maps. May anyone can give me an idea to solve this problem? Sorry if my question trouble you because I'm newbie in both Matlab and Computer Vision.
Thank you so much.