AMBIENT LIGHTING DERIVED FROM VIDEO CONTENT USING ADAPTIVE EXTRACTION REGIONS

摘要

Extracting video content encoded in a rendered color space for broadcast by an ambient light source, using adaptive extraction regions that can change their size, shape, location, and distribution. Perceptual rules in concert with user preferences can provide for intelligent dominant color selection. Steps include quantizing the video color space; performing dominant color extraction from adaptive extraction regions by using a mode, median, mean, or weighted average of pixel chromaticities; and applying perceptual rules to further derive dominant chromaticities via [1] chromaticity transforms; [2] a weighted average using a pixel weighting function influenced by scene content; [3] extended dominant color extraction where pixel weighting is reduced for majority pixels. Adapting the extraction regions allow for changing the luminance, the chromaticity, the temporal delivery, or the spatial extraction of said dominant color.