Optimal Block Cosine Transform Image Coding for Noisy Channels

Vinay A. Vaishampayan, Nariman Farvardin

Research output: Contribution to journalArticlepeer-review

37 Scopus citations

Abstract

The two-dimensional block transform coding scheme based on the discrete cosine transform has been studied extensively for image coding applications. While this scheme has proven to be efficient in the absence of channel errors, its performance degrades rapidly over noisy channels. In this paper, we present a method for the joint source-channel coding optimization of a scheme based on the 2-1) block cosine transform when the output of the encoder is to be transmitted via a memory less binary symmetric channel. Our approach involves an iterative algorithm for the design of the quantizers (in the presence of channel errors) used for encoding the transform coefficients. This algorithm produces a set of locally optimum (in the mean-squared error sense) quantizers and the corresponding binary codeword assignment for the assumed transform coefficient statistics. To determine the optimum bit assignment among the transform coefficients, we have used an algorithm based on the steepest descent method, which under certain convexity conditions on the performance of the channel-optimized quantizers, yields the optimal bit allocation. Simulation results for the performance of this locally optimum system over noisy channels have been obtained and appropriate comparisons against a reference system designed for no channel errors have been rendered. It is shown that substantial performance improvements can be obtained by using this scheme. Furthermore, theoretically predicted results and rate distortion-theoretic bounds for an assumed 2-D image model are provided.

Original languageEnglish
Pages (from-to)327-336
Number of pages10
JournalIEEE Transactions on Communications
Volume38
Issue number3
DOIs
StatePublished - Mar 1990

Fingerprint

Dive into the research topics of 'Optimal Block Cosine Transform Image Coding for Noisy Channels'. Together they form a unique fingerprint.

Cite this