Robust skin detection based on the fuzzy integral

Skin detection is a low-level operation used in multiple application fields, e.g. video surveillance, human computer interface. The paper presents a multi-procedure methodology followed by a fu- sion stage that can be applied in the robust detec- tion of skin. Two well-known methodologies are fused with a novel one, which is presented here as well, through the application of the fuzzy integral. The automated parameterization of the resulting procedure and its performance on a benchmark database are finally described.