Secondary quantization of color images for minimum visual distortion

摘要: Many times in electronic imaging systems it is necessary to reduce the precision of digital data for display, storage, manipulation, or transformation an image. For example, may be 12 bits/channel RGB 8 due storage requirements. To accomplish this reduction number levels, a series input levels must grouped together each output level. Since process involves quantization previously quantized data, sometimes referred as secondary quantization. The necessarily results artifacts, such contouring image, where many colors have been mapped single color. Conventional methods, linear power-law resampling, are suboptimal and do not consider intercolor effects. This paper describes method determining functions that will minimize observable image artifacts generated by process. basic approach use nonlinear optimization techniques cost function provides measure visible color error. Examples presented compare optimized conventional techniques.© (1993) COPYRIGHT SPIE--The International Society Optical Engineering. Downloading abstract permitted personal only.