Features Extraction for Robust Face Recognition Using GLCM and CS-LBP

Abstract

For fast and efficient automatic identity verification, biometric technology has developed rapidly. As one of the useful biometric identification technologies is face recognition. The applications of the face identification system-wide such as criminal detection, security verification, credit card verification, medicine, video conference, and other occasions. The main elements of the pattern recognition system include preprocessing, feature extraction, and classification. The feature extraction stage is concerned in this paper in order to suggest a robust algorithm useful in the classification stage. Many types of research try to recognize the face utilizing various algorithms, comparison among these algorithms is presented in this paper to show the effect of the feature extraction algorithm on human face classification. The main goal of this research is to combine Gray Level Co-Occurrence Matrix (GLCM) and Center-Symmetric Local Binary Pattern (CS-LBP) in a manner way to be used for extracting the texture feature of the human face classification. The experimental result explains that the proposed method given high accuracy in recognizing the human face where the proposed algorithm capable of recognizing 28 humans under different poses and illumination conditions.

A Novel Approach for Object Recognition Using Decision Tree Clustering by Incorporating Multi-Level BPNN Classifiers and Hybrid Texture Features

“…We built our object recognition system using these two unique powers of the fractal parameter. Also, in view of wide applications of GLCM in image classification problems (Salman et al, 2022;Vera et al, 2023;Alazawi et al 2019), GLCM feature was also computed to build the hybrid texture feature.…”

“…In order to prepare hybrid feature, other texture feature Gray-level Co-occurrence Matrix (GLCM) was explored in this work. Many researchers have used GLCM as a texture feature in facial recognition process (Salman et al, 2022;Vera et al, 2023;Alazawi et al 2019). Finally hybrid feature was prepared by combing ILMFD+GLCM in order to train the model effectively.…”

Generating and Detecting Face Morphing Using Texture Techniques

“…This is done through specific angles 45, 90, 180, 270 and then applying statistical measures to this matrix. Figure 4 shows calculate of GLCM [14,15].…”