Abstract: In order to recognize a face, many processes contribute to complete this process such as detection of a face, image processing, feature extraction, and classification. All these processes are executed in different ways. There are several techniques to carry out these tasks. Researchers have delved deeper into this field to study different techniques and their efficiency in terms of recognizing the face. A significant number of studies have been carried out to this end. Recently, a technique of feature extraction and face recognition algorithm was proposed in which PCA was used to extract the features: which is efficient but lacks in some aspects. Thus, in this paper, a novel approach is presented in which the concept of Region of Interest (RoI) is introduced to the input image. PCA feature extraction is replaced by implementing hybrid LBP-LPQ. These techniques offer various advantages over PCA. Multi-class SVM is used for classification purposes. MATLAB is used to perform simulation results. Performance evaluation of the proposed method is carried out in terms of different parameters such as recognition time for different numbers of samples and recognition rate for different dimensions and number of samples. From the comparative analysis, the proposed technique outperformed the existing approaches.
Keywords: Face recognition, Region of interest, LBP, LPQ
|
DOI:
10.17148/IJARCCE.2020.9218
