A Novel Approach towards Facial Skin Detection using Back propagation Neural Network

Abstract

Detecting faces in an image is a prime step in various computer vision applications, such as face recognition, facial expression analysis etc. Facial skin detection is the process of finding skin-colored pixels and regions in an image with respect to other skin present in the body. Several computer vision approaches like Knowledge-based, Feature invariant approaches, Template matching, Appearance-based methods have been developed for skin detection. A neural network based facial skin detector typically transforms a given pixel into an appropriate colour space and there, the work of the neural network based classifier is to label the pixel whether it is a skin or a non-skin (other than facial skin) pixel. A facial skin classifier defines a decision boundary of the skin color class in the colour space based on a training database of skin-colour pixels. In this paper we have used a neural network based classifier to identify a skin pixel from a non-skin pixel by using a back propagation neural network. Non-skin here refers to the skin other than the facial skin.