Vector quantizing method for usage on channels with limited bitrate

摘要

The method involves determining from the points of the network in a predetermined region of each quantised point the best compromise in terms of a predetermined relationship between the distortion arising from the quantisation of the point and the bit rate arising from the length of the code allocated to that point. Those corresponding to the determined points are retained as quantised points for coding. The quantised points can belong to several networks and the distortion is minimised among them. In both cases, the polar coordinates of each quantised point are coded. A variation minimises (d + Lr) where d is the distortion, r is the bit rate obtained and L is the Lagrange constant.