Klasifikasi Kualitas Garam Tambak Menggunakan Metode Gray Level Co-occurrence Matrix (GLCM) dan K-means

Abstract

The amount of salt production in Rembang Regency in the last four years has continued to decline. The results of this study are expected to be a consideration in improving the quality of salt production in Rembang Regency. Farm salt is classified into three salt qualities, namely K1 quality, K2 quality, and K3 quality. The difference in color at each level of salt quality is a problem for some people due to the lack of knowledge about the level of color brightness for each salt quality. This research uses the GLCM method for salt feature extraction and classification using the K-means algorithm. The data used in this study are extraction data and testing data. The results of feature extraction on training data obtained the average value of the GLCM features of the data for each quality used as the initial centroid. Quality K1 obtained contrast value: 0.0387 energy: 0.8039 homogeneity: 0.9823 entropy: 0,6770. Quality K2 is obtained from the contrast value: 0.1361 energy: 0.4387 homogeneity: 0.9365 entropy: 1,8904. K3 quality is obtained from the contrast value: 0.1920 energy: 0.3123 homogeneity: 0.9108 entropy: 2,3881. The testing results of salt using geomembrane have an average quality that is better than salt without geomembrane. The results based on subdistricts in Rembang sub-district obtained K1 quality as much as 8, K2 as much as 1, and K3 as much as 1. Lasem sub-district obtained K1 quality as much as 1, K2 as much as 6, and K3 as much as 2. Sarang sub-district obtained K2 quality as much as 2, and K3 as uch as 2. Sluke sub-district obtained K1 quality as much as 5, K2 as much as 1, and K3 as much as 1. Kaliori sub-district obtained K1 quality as much as 5, K2 as much as 2, and K3 as much as 3.