4G Wireless Video Communications

3

Video Coding and Communications

3.1 Digital Video Compression – Why and How Much?

Uncompressed standard definition (SD) video, captured by following the ITU-R 601 4:2:2 standard [1] with 10 bits per pixel, consists of a 270 Mbits/sec bit rate. The typical channel capacity that is available and is used for the transmission of a TV channel is less than 6 Mbits/sec and averages around 3 Mbits/sec. There is a strong push to further reduce that capacity due to economic reasons. This implies that one is required to remove more than 98% of the original information and compress the digital video to less than 2% of the initial bit rate. It requires more than 50 × compression factor. It is not possible to achieve that amount of compression without loss of information. The loss of information adds distortion to the decoded video which appears as a degradation of visual quality in the displayed video. Therefore, one of the key goals of a compression algorithm is, at a given bit rate, to add as little distortion in the decoded video as possible and to lose the information so that the degradation is as invisible as possible. To minimize the perceptual distortion some processing steps are taken before encoding the video. The typical pre-compression steps taken for consumer applications are to compress only the visible part of the video, reduce the number of bits per pixel from 10 to 8 bits and reduce the color resolution before compressing the video.

The visible part of a frame of video contains ...