Human-Face Image Retrieving Based Texture Feature Extraction Method

The face image retrieval solutions are widely studied, Although it deal with different facial features and difficulties to retrieve facial images due to the similarity of these features. This paper introduces a solution of human face-images retrieval based on a combination feature extraction methods where there are some challenges related with human-face image detection and then retrieving. Human face-images retrieving used in various fields like justification, Criminal Evidence and law inspections and robotic intelligence. The research-proposal contributes to conquer on a number of confronts in images of human-face detections then accurate retrieving in acceptable time. So this work deals with getting higher rate of recognition using a combination method of Gray-Level-Co-occurrence-Matrix(GLCM) and Local-Binary-Patterns (LBP) as feature-descriptors and classifiers to develop face-image recognition. First of all some previous processing techniques of image to detect the centre of face image then GLCM calculation method involves process of gray image, after that a number of statistical-texture attributes and 2nd order-attributes are obtained. The LBP technique acts as feature extraction after the representation of a human-face image, and finally the classification. The histograms are finding of blocks of an image of human-face. Then retrieve human-face image based minimum difference between attributes of a strange human-face image with the features of familiar images. All findings of this work were evaluate using MSE, Chi-square test and PSNR. Olivetti Research Laboratory ORL human-face images dataset used in this proposal. The experiments showed that the combination technique detecting the human-faces grows accuracy rate, and effectiveness. The results show increase in recognition exactitude to be 98%.