Skin detection using neighborhood information

Abstract

Skin detection is employed in tasks like face detection and tracking, naked people detection, hand detection and tracking, people retrieval in databases and Internet, etc. However, skin detection is not robust enough for dealing with some real-world conditions, like changing lighting conditions and complex background containing surfaces and objects with skin-like colors. This situation can be improved by incorporating context information in the skin detection process. For this reason in this article a skin detection approach that uses neighborhood information is proposed. A pixel will belong to the skin class only if a direct neighbor does. This idea is implemented through a diffusion process. Two new algorithms implementing these ideas are described and compared with state-of-the-art skin detection algorithms.