Open AccessIdentification of Tuna and Skipjack Fish Texture Using GLCM With Naive Bayes Classification
Author(s) -
Muzakir Hi. Sultan,
Ruslan Laisouw
Publication year - 2020
Publication title -
agrikan
Language(s) - English
Resource type - Journals
eISSN - 2621-0193
pISSN - 1979-6072
DOI - 10.29239/j.agrikan.13.2.285-291
Subject(s) - tuna , skipjack tuna , artificial intelligence , pattern recognition (psychology) , naive bayes classifier , mathematics , scombridae , fish <actinopterygii> , statistics , computer science , biology , fishery , support vector machine
Fish in Indonesian waters have various types, the famous ones are tuna and skipjack. The two types of fish look similar, because they come from the same family, namely scombridae. To find out and differentiate types of tuna and skipjack fish, it can be seen based on the texture image. The method that can be used in analyzing texture is the Gray Level Coocurent Matrix (GLCM) method. There are several methods of image classification, one of which is the Naive Bayes method. This study aims to identify types of tuna and skipjack based on texture analysis using the GLCM and Naive Bayes methods. Based on the results of testing data analysis on types of tuna and skipjack meat using GLCM with angles and , the distance of each pixel is 1, indicating the value of Energy, Entropy Contrast, Homogeneity, Correlation, Sum Average, and Sum of Variance are highly varies. As well as the Naive Bayes classification results obtained a probability of 0.58 or 58% categorized as tuna meat, while the remaining probability of 0.42 or 42% is categorized as skipjack.