作者: Y. Wu , D.C. Coll
DOI: 10.1109/26.231961
关键词: Lossless compression 、 Color Cell Compression 、 Vector quantization 、 Data compression 、 Mathematics 、 Pattern recognition 、 Lossy compression 、 Data compression ratio 、 Block Truncation Coding 、 Artificial intelligence 、 Context-adaptive binary arithmetic coding
摘要: An encoding technique called multilevel block truncation coding that preserves the spatial details in digital images while achieving a reasonable compression ratio is described. adaptive quantizer-level allocation scheme which minimizes maximum quantization error each and substantially reduces computational complexity of optimal levels introduced. A 3.2:1 can be achieved by itself. The truncated, or requantized, data are further compressed second pass using combined predictive coding, entropy vector quantization. lossless lossy. total ratios about 4.1:1 for second-pass compression, 6.2:1 lossy compression. subjective results algorithm quite satisfactory, with no perceived visual degradation. >