In this paper there is enclosed a description of the hybrid system for black and white (256 by 256 pixels) images compression. The system includes the following procedures: image decomposition (8x8, 16x16 blocks), DCT transformation, “zig-zag” scanning, product code vector quantization (one-dimensional block) and a bit allocation. The standard LBG algorithm for codebook design has been enriched with the simulated annealing procedure for avoiding the local minima. Standard vector quantization in two-dimensional transform space has been investigated for comparison.
CITATION STYLE
Rak, R. J. (1998). Transform vector quantization of images in one dimension. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1424, pp. 361–368). Springer Verlag. https://doi.org/10.1007/3-540-69115-4_49
Mendeley helps you to discover research relevant for your work.