作者: R.Y. Wang , E.A. Riskin , R. Ladner
DOI: 10.1109/ICASSP.1993.319790
关键词: Linde–Buzo–Gray algorithm 、 Communication channel 、 Error detection and correction 、 Mathematics 、 Pattern recognition 、 Maximum a posteriori estimation 、 Vector quantization 、 Artificial intelligence 、 Bit error rate 、 Codebook 、 Decoding methods
摘要: The authors describe a new way to organize full search vector quantization codebook so that images encoded with it can be sent progressively and have immunity against channel noise. Due the organization, most significant bits (MSBs) of codeword index are important overall image quality highly correlated. Simulations show effective error rates MSBs substantially lowered by implementing maximum posteriori (MAP) detector similar one suggested N. Phamdo Farvardin (1992). performance scheme is close pseudo-Gray coding at low bit outperforms higher rates. >