作者: Kwok-tung LO , Wai-kuen CHAM , None
DOI: 10.1016/B978-0-444-89587-5.50011-6
关键词: Linde–Buzo–Gray algorithm 、 Process (computing) 、 Encoder 、 Codebook 、 Artificial intelligence 、 Vector quantization 、 Mathematics 、 Encoding (memory) 、 Pattern recognition 、 Code word 、 Distortion
摘要: ABSTRACT: In vector quantization of images, the main problem is its computation complexity in encoding process. this work, a sub-codebook searching (SCS) algorithm developed for fast VQ images. This allows only portion codebook to find minimum distortion codeword an input vector. comparison with other existing algorithms, method requires least number multiplications as well total operations. The requirement extra memory at encoder and pre-computation training stage SCS also limited.