A Method of Thai Main Dish and Soup Classification by Gray Level Co-Occurrence Matrix Algorithm

Abstract

This paper presents a new method to classify Thai main dish and soup by Gray Level Co-occurrence Matrix (GLCM). An entropy-based algorithm is used to extract the food area from the background. A GLCM texture analysis algorithm is used to calculate features vector of food. The GLCM algorithm can calculate an energy, a homogeneity, and a correlation to be featured. The three parameters are used for classification. Support Vector Machine technique is used for classification. The experimental result showed 100 percent of sensitivity and specificity for main courses, 89.9 percent of sensitivity and specificity and 99.44 percent of accuracy. This is the first method that can classify Thai main dish and soup. © 2018 IEEE.