Indonesian Ethnicity Recognition Based on Face Image Using Uniform Local Binary Pattern (ULBP) and Color Histogram

Abstract

Ethnicity is one of identity every human has and can be used to categorize individuals in populations or large groups. We presented an Indonesian ethnicity recognition based on facial images using Uniform Local Binary Pattern (ULBP) and Color Histogram as a feature extraction method. We used the five largest ethnic groups in Indonesia, namely Sundanese, Javanese, Banjar, Buginese, and Malay. In the experiment, we used Random Forest as a classification method. The research obtained a performance accuracy of 98.25% using 2290 facial images.