F5 algorithm is available on github for jpeg image. You have to re-implement it for video instead of LSB substituition, with some more functionalities(see attached file). GUI would be nice and any programming language can be used. You have to make document/comments to make the code and algorithm understandable to me. Have to get some steganalysis results of the stego-work and find out a way to compare with LSB video steganography (available online).