Face Recognition Using Morphological Analysis of Images
Face recognition from still and motion image has been an active and emerging research area in the field of image processing, pattern recognition and so on in the recent years . The challenges associated with discriminant face recognition can be attributed to the following factors such as pose, facial expression, occlusion, image orientation, image condition, presence or absence of structural component and many more. In this paper, we have tried to emphasize on the morphological analysis of images based on the behavior of the intensity value. Firstly images with various situations of a person are selected as training images. Based on the min, max and average characteristics of images, the training model has been built. Morphological analysis like binary image processing, erosion and dilation play the important role to identify the facial portion of an image from the whole one. Finally face recognition has been made for input images based on their intensity value measurement. The training images collected from various database such as YALE, ORL, and UMIST and others. The algorithm performed well and showed 80 percent accuracy on face prediction
