I watermarked a picture in sound .WAV file, when today i find this sound file after 3 years try to extract picture embed in sound but i understand format of this chang to .mp3 and thus I can't extract completely picture of watermarked .My question about , find a watermarking algorithm that is resistant to any kind of compression?
Watermarking and steganography are very related to each other. While steganography is interested in capacity and invisibility, watermarking is to provide robustness to protect hidden data against any attacking methods like compressing, cropping, noise addition etc. LSB Coding, Phase Coding, Spread Spectrum, Echo Hiding are basic methods you should search about.
I suggest you to read these articles for the basics:
First article is very helpful to start. I advice you to try Spread Spectrum method, it was very robust against mp3 compression as I have coded myself. I had answered my own question here trying to explain steps of Direct Sequence Spread Spectrum Method to embed data inside audio files. It may also help you.