Morphological based Face Detection & Recognition with Principal Component Analysis

This paper includes face detection and recognition with the help of morphological shared weight neural network. Face detection is achieved with the combination of morphological hit miss operation and edge detection. Feature extraction of face image is done using Principal Component Analysis. These features are used to train the neural network. The output of neural network is compared with database and depending on that comparison recognition is achieved. In neural network, numbers of input nodes used are varying depending on number of persons & output node is 1 with single hidden layer.. The experimental results on IIT Kanpur Indian database shows good reliability and performance with testing accuracy of 82.50%, so it is promising to be used in a personal identification system. Keywords— Edge Detection, Hit miss transform, Neural network, PCA.