Skin Color Segmentation in RGB Color Space by Adaptive Network Based Fuzzy Inference System (ANFIS)

Skin color detection is a popular and useful technique because of the wide range of application in both human computer interactions and analyses based on diagnostic. Therefore, providing an appropriate method for pixel-like skin parts can solve many problems. The presented color segmentation algorithm works directly in RGB color space without having to convert the color space. Using Genfis3 function, we formed the Sugeno fuzzy network and clustered the data using fuzzy C-Mean (FCM) clustering rule and for each class and cluster we considered a Rule. In the next step, the output resulting from pseudo-polynomial data mapping is provided as the input to Adaptive Network Based Fuzzy Inference System (ANFIS).