Classification of beef and pork meat by using SVM algorithm

Abstract

Meat is the main parameter in human daily consumption, because the excess protein in meat can increase human intelligence and stamina, but the similarity in color and texture of two types of meat makes it difficult for the human eye to distinguish them. This is used as a loophole by irresponsible meat traders to mix two types of meat, so it can cause serious health conditions in humans. Meat can generally be distinguished by looking at the fibers and colors of certain meats. Therefore, this study will overcome this problem by prioritizing the parameters of different colors and textures of meat which are classified as optimal methods with digital image processing. Among the efficient approaches offered is to apply the support vector machine (SVM) method by extracting HSV features in distinguishing beef and pork categories. This study adopts SVM and HSV feature extraction in finding out the types of beef and pork with fairly good algorithm results of 97% in 510 images, which were divided into training and testing data. The dataset used in the study is taken from Kaggle as an open-source platform for any practitioner. As a result, this technology shows a very accurate opportunity in detecting the type of meat correctly.