Face identification using soft computing tool
Debaraj Rana, Nrusingha Prasad Rath · 2012
In this paper we have proposed a novel approach for face identification as application of Genetic Algorithm, based on PCA and applicable to colour images. Our proposed method is a search procedure to find out the face of interest from a 2D image. For optimization of search procedure we have used here Genetic algorithm with a fitness function. The proposed method is based on Eigenfaces method using PCA and a term distance from face space (DFFS) which is a Euclidian distance that is used here in fitness function for GA. And for quicker convergence we have first extracted the skin area from the scene image from YCbCr colour model. This method has specific application like locating a criminal from a scene image. This method is very fast due to reduction in search space.