Applying neuron network to improve characteristic of video encoder

Abstract

The paper deals with h.263 recommendation for the video stream compression. Purpose of the modification is a stream compression enhancement with insignificant losses in picture quality. Enhancement of the video stream compression is achieved by unsupervised artificial neural network. The artificial neural network is used for generating codebook prototypes for blocks with over-average level of changes. Dynamic video sequence from Matrix movie is used in purpose of comparative testing of the described enhanced h.263 video codes. © 2008 IEEE.