Chromatic colour spaces for skin detection using GMMS

Abstract

Skin detection often forms part of both face detectors and pornography detectors for Internet filters. This paper introduces a trivial skin detection algorithm that uses Gaussian Mixture Models to classify 8 × 8 blocks according to their skin-likeness. The absence of structural features in skin regions means luminance information alone is unsuitable for accurate skin detection. Therefore, the affect of the chosen colour components on the classification accuracy is investigated. Fused features vectors extracted from the Cb and Cr chrominance components were found to give the best results, achieving an equal error rate of 5.5%. By utilising these components and an 8 × 8 block size, skin regions in JPEG and MPEG files can be located without full decompression.