A Comparison of Wavelet Coding Sequences for Noisy Channels

A new wavelet-based image compression scheme is proposed for use on noisy communications channels. In variable-length coding schemes, the occurrence of a single channel error can result in the entire image being lost. We adopt a simple approach based on error detection and recovery to stem the propagation of errors in the decoding process. Once the combined source/channel coding method is deened, we shift our attention to the sequence in which wavelet coeecients are encoded. We describe four diierent coding sequences and perform experiments to discover if the coding sequence has a signiicant impact on reconstructed image quality, and, if so, which sequence is best for noisy channels.