Identifikasi Tingkat Kematangan Buah Pinang Menggunakan K-Nearest Neighbor Berdasarkan Fitur Tekstur dan Warna

Abstract

This research builds a system for identifying the maturity level of areca fruit based on digital image processing using texture and color features through the Gray Level Co-Occurrence Matrix (GLCM) and Color moments. The initial stage of the research is image pre-processing so that it can be processed to the next stage, namely feature extraction. Texture feature extraction was performed using the Gray Level Co-Occurrence Matrix (GLCM), namely the correlation value and color feature extraction using Color moments, the mean value used in this study. Classification is done based on the features that have been extracted before. This study uses the K-Nearest Neighbor (KNN) classification method. Tests were carried out to determine the parameters that cause changes in the classification results with scenarios including determining the number of Neighbors in KNN. By using 1 Neighbors in the KNN classifier, the best accuracy is 86.36% in the process of identifying the maturity level of areca fruit.