tailieunhanh - Face detection and recognition is challenging due to the Wide variety of faces and radial basis function network

In this paper, we propose a neural network based novel method for face recognition in cluttered and noisy images. We use a Modified radial basis function network (RBFN) to distinguish between face patterns and non face patterns. | ISSN:2249-5789 KesavaRao Seerapu et al , International Journal of Computer Science & Communication Networks,Vol 2(5), 584-589 KesavaRao Seerapu1 R. Srinivas2 1 ( Student, Dept. of CSE, Aditya Institute of Technology and Management, Tekkali, Andhra Pradesh 2 (Associate Professor, Dept. of CSE, Aditya Institute of Technology and Management, Tekkali, Andhra Pradesh Abstract— Face detection and recognition is challenging due to the Wide variety of faces and the complexity of noises and image backgrounds. In this paper, we propose a neural network based novel method for face recognition in cluttered and noisy images. We use a Modified radial basis function network (RBFN) to distinguish between face patterns and non face patterns. The complexity RBFN is reduced by RobustPCA as it gives good results even in different illumination environments and highly un-susceptible to occlusion when compared with Classical PCA (Principal component analysis). RobustPCA is applied on Images to get the eigen-vectors. These eigen-vectors are given as input to RBFN network as the inputs for training and recognition. The proposed method has good performance good recognition rate. Keywords: RobustPCA, Neural networks, Radial basis function network, Face recognition, Eigen vectors, PCA method for face recognition using RobustPCA and RBFN. These systems can be well incorporated into mobile and embedded systems efficiently and can be utilized on larger recognition becomes challenging with varied illumination and pose conditions. This method over comes the varied illumination problem and detection in noisy environments. 2. System Overview The procedure for Face recognition is as follows. 1. Pre processing: The image is rescaled and the noise is reduced, contrast was normalized with histogram equalization 2. RobustPCA: The images then are applied with RobustPCA for reduction in dimensionality and there by reducing complexity. 3. Modified RBFN: The outputs of .