Transform coding of images using trellis coded quantization

Trellis-coded quantization (TCQ) is incorporated into a transform coding structure for encoding monochrome images. Both fixed-rate and entropy-constrained designs are considered. The fixed-rate TCQ-based systems provide gains in peak signal-to-noise ratio of up to 4.13 dB over scalar quantizer-based schemes, while the entropy-constrained TCQ-based systems provide gains of up to 6.39 dB. The subjective quality of the TCQ-based systems is substantially better than that of comparable scalar quantizer-based systems. The high-frequency background noise, fuzziness, and block artifacts produced by the scalar quantizer-based systems are virtually eliminated by the TCQ-based systems.<
>