A LOW-COMPLEXITY PREDICTED VECTOR QUANTIZATION SCHEME

Recently, the use of spatial prediction schemes had been adopted in Ve ctor Quantization for enhancing the coding fidelity and visual quality in image compression. Despite the improvement, these schemes are gene rally computationally expensive and hence increase the complexity of t he encoders and decoders. In this paper, two Predicted Vector Quantiza tion techniques had been reported. The proposed methods are economical and simple to implement, and provide reasonable improvement over the classical LBG algorithm.