An Adaptive Video Coder using Saliency and Second Generation Wavelets

S. Brangoulo and P. Gioia (France)


Video and Picture Coding, Second Generation Wavelets,Hierarchical Mesh Representation.


We propose a method based on second generation wavelets. After having tested the first generation wavelets hybridised with the latest video standard (the DCT has been replaced by the wavelets), we have shown that this transform is not adapted to our particular data. The new wavelet transform is performed using the powerful lifting scheme, keeping the main properties of the first generation wavelets whilst introducing some useful properties such as: analysis on general domains, suitable basis for weighted approximation / interpolation and weighted measure and transform adapted to irregular sampled data. With second generation wavelets, one can locate and encode the residual error of the motion compensation process, which mainly consists of high frequencies and singularities, typically the kind of data that second generation wavelet can successfully analyse.

Important Links:

Go Back