CONSTRAINED-STORAGE VECTOR QUANTIZATION WITH A UNIVERSAL CODEBOOK

Many image compression techniques require the quantization of multiple vector sources with significantly different distributions. With vecto r quantization (VQ), these sources are optimally quantized using separ ate codebooks, which may collectively require an enormous memory space , Since storage is limited in most applications, a convenient may to g racefully trade between performance and storage is needed, Earlier wor k addressed this problem by clustering the multiple sources into a sma ll number of source groups, where each group shares a codebook, We pro pose a new solution based on a size-limited universal codebook that ca n be viewed as the union of overlapping source codebooks, This framewo rk allows each source codebook to consist of any desired subset of the universal codevectors and provides greater design flexibility which i mproves the storage-constrained performance. A key feature of this app roach is that no two sources need be encoded at the same rate. An addi tional advantage of the proposed method is its close relation to unive rsal, adaptive, finite-state and classified quantization, Necessary co nditions for optimality of the universal codebook and the extracted so urce codebooks are derived. An iterative design algorithm is introduce d to obtain a solution satisfying these conditions. Possible applicati ons of the proposed technique are enumerated, and its effectiveness is illustrated for coding of images using finite-state vector quantizati on, multistage vector quantization, and tree-structured vector quantiz ation.