Implementasi Learning Vektor Quantization (LVQ) dalam Mengidentifikasi Citra Daging Babi dan Daging Sapi

Abstract

Widespread circulation of adulterated meat and based on the word of Allah which confirms the prohibition of pork to eat, it needs to be made of a system that can distinguish between beef and pork to avoid cheating merchants and keep halal meat we eat. This study makes a system for identifying the image of beef and pork and meat adulterated with the color feature extraction HSV (Hue, Saturation, Value) and texture feature extraction GLCM (Grey Level Co-occurent Matrix) using classification LVQ (Learning Vector Quantization). A result of image identification adulterated meat pig is considered as a pork class. Image data on the image of the study consisted of 107 primary and 13 secondary image. Identification testing conducted on the distribution of training data and test data are different. Accuracy of the highest success with an average of 94.81% on the distribution of the 80 training data and test data 20 and the accuracy of the lowest success with an average of 82.22% on the distribution of training data and test data 50 50 with Learning Rate of 0.01, 0.05, 0.09. More increase the distribution of training data and more decrease division of the test data, so more increase the accuracy of success in identifying the image.
Keywords: beef, GLCM, HSV, Learning Rate, LVQ, pork