作者: Eve A. Riskin , Vaughn S. Iverson
DOI:
关键词: Personal computer 、 Palette (computing) 、 Mathematics 、 Workstation 、 Image quality 、 Distortion 、 Artificial intelligence 、 k-nearest neighbors algorithm 、 Computer vision 、 Image (mathematics) 、 Pairwise comparison
摘要: A color-mapped display sub,stem efficiently combines palettes of multiple images into a single shared palette. As each image already received degree distortion during conventional palette selection, it is desirable to minimize further the combination method this invention. pairwise nearest neighbor (PNN) technique used for combining colors from respective distortion. For final 256-color palette, up 256 (n-1) individual vector merges are performed (where n number being combined). In one embodiment, two vectors chosen at step that yield lowest increase in when merged. mean squared error measure gamma-corrected values defined YIQ space compare Searching time reduced O(N 2 ) O(N), while also eliminating need extensive recalculation color pair distortions between steps. Efficiency enhanced because matrix effectively caches calculations One advantage invention ability service run-time demands simultaneous on personal computer or workstation platform having an 8-bit subsystem. Another maintenance quality across similar using