Texture-based Feature using Multi-blocks Gray Level Co-occurrence Matrix for Ethnicity Identification

Abstract

Ethnicity identification for demographic information has been studied for soft biometric analysis, and it is essential for human identification and verification. Ethnicity identification remains popular and receives attention in a recent year especially in automatic demographic information. Unfortunately, ethnicity identification in a multi-class which consist of several ethnic classes may degrade the accuracy of the ethnic identification. Thus, this paper purposely analyses the accuracy of the texture-based ethnicity identification model from facial components under four-class ethnics. The proposed model involved several phases such as face detection, feature selection, and classification. The detected face then exploited by three proposed face block which are 11, 12 and 22. In the feature extraction process, a Grey Level Co-occurrence Matrix (GLCM) under different face blocks were employed. Then, final stage was undergone with several classification algorithms such as Naïve Bayes, BayesNet, k-Nearest Neighbour (k-NN), Random Forest, and Multilayer Perceptron (MLP). From the experimental result, we achieved a better result 22 face block feature compared to 11 and 22 feature representation under Random Forest algorithm.