Abstract

Facial recognition has emerged as the most promising and robust method of recognizing people in recent years. The research focuses on performance assessment of Local Binary Patterns (LBP), and Chicken Swarm Optimization (CSO) face recognition techniques. The Local Binary Patterns (LBP) technique and Enhanced LBP were used to extract features, Chicken Swarm Optimization (CSO) algorithm and Improved CSO were used for feature selection, and the Support Vector Machine (SVM) was used as a classifier. Performance assessment was done by comparing the combination of techniques LBP-CSO, CSO-ELBP, ICSO-LBP, and ICSO-ELBP. Experimental results in terms of recognition time and recognition accuracy were employed, using the KWASU database. The results show that the LBP-CSO has an accuracy of 91.67% at 119.10 seconds, ELBP-CSO has an accuracy of 95.00% at 79.16 seconds, and LBP-ICSO has an accuracy of 96.25% at 105.20 seconds and ELBP-ICSO has an accuracy of 97.92% at 58.37 seconds.