In this paper, Face recognition based on clustering feature extraction using gaussian mixture model has been implemented. Face recognition system is widely used in security purpose. The preprocessing of the face recognition system is used to extract the features to classify the test image from the large set of face databases. On behalf of face recognition, most of the feature extractions are based on the various transforms such as decomposition of the multidimensional wavelets. The disadvantages of the wavelets are it does not cover the spatial variations of the training face images. Hence, the machine learning method very popular tool to recognition face image. Gaussian mixture model is an unsupervised learning based on the clustering process to extract the features based on expectation-maximization method. The standard databases are ORL, Yale, and real-time database have implemented. The performance measures are Precision, Recall and F-Score have been verified. The proposed method brings improved results over existing face recognition methods.
Volume 11 | 07-Special Issue
Pages: 1626-1635