Klasifikasi Jenis Dan Tingkat Kesegaran Daging Berdasarkan Warna, Tekstur Dan Invariant Moment Menggunakan Klasifikasi LDA

Classification of Type and Freshness Level of Meat Based on Color, Texture and Invariant Moment Using LDA Classification

Abstract

Distribution process that takes a long time along with improper treatment, can cause meat become not fresh and decrease the quality of the meat. Therefore, unscrupulous meat sellers cheating on the non fresh meat by mixing the non fresh meat with the fresh one. A system that can classify the type and freshness level of meat automatically is needed. In this research, that system was developed based on texture, color and shape features using Linear Discriminant Analysis (LDA) classification. The methods used in the feature extraction process are statistical approach, GLCM and the HU's invariant moment. The total of data used in this research was 960 images of 3 different meat types which are chicken meat, goat meat, and beef. The highest accuracy obtained from the testing process was 90% on the combination features of HSI and invariant moment for the meat type in refrigerator.