Robust quantization for image coding and noisy digital transmission

Abstract

A robust quantizer is developed for encoding a variety of memorylesssources and transmission over the binary symmetric channel (BSC). The system combines channel optimized scalar quantization (COSQ) with all-pass filtering, the latter performed using a binary phase-scrambling/descrambling method. Applied to a broad class ofsources, the robust quantizer achieves the same performance as the Gaussian COSQ for the memoryless Gaussian source. This quantizer is used for image coding for transmission over a BSC. An explicit error protection code is used only to protect side information. The PSNR performance degrades gracefully as the channel bit error rate increases. I IntroductionApproaches to designing scalar and vector quantizers for use over a binary symmetric channel (BSC) include Farvardin and Vaishampayan's channel optimized scalar quantization (COSQ) [l], and channel optimized vector quantization (COVQ) [2], Ayanogu and Gray's trellis waveform coding [3], and Wang and Fischer's channel optimized trellis coded quantization (COTCQ) [4]. For a given VQ design, Farvardin [5], Zeger and Gersho [6], and Hagen and Hedelin [7] have developed algorithms to assign binary codewords to index the VQ codevectors to minimize the distortion caused by channel errors. This has been called pseudo-Gray VQ [6]. Laroia and Farvardin's scalar vector quantizer (SVQ) [8] is representative of the good VQ performance that can be achieved, provided channel errors are not a significant problem. Figure 1 compares the performance of these methods for a memoryless Gaussian source. It is evident that if the channel bit error rate is significant (larger than about lo-') very