Pengaruh Ekstraksi Fitur Tekstur Pada Hasil Klastering Data Citra Buah Menggunakan Metode K-Means Cluster

Veronica, Lusiana and Imam Husni, Al Amin and Budi, Hartono Pengaruh Ekstraksi Fitur Tekstur Pada Hasil Klastering Data Citra Buah Menggunakan Metode K-Means Cluster. Journal of Computer System and Informatics (JoSYC).

PDF Download (1MB)

Abstract

This study aims to analyze the effect of texture feature extraction using the grey level co-occurrence matrix (GLCM) and local binary pattern (LBP) methods on the clustering results of image data. Texture feature extraction is performed on test data of 30 ripe fruit images and rotten fruit images. Through experiments, it was found that the LBP feature extraction method can increase the value of contrast features and decrease the value of correlation features. In the energy feature, with or without LBP, the difference in the value of this feature is not too far. The GLCM and LBP methods affect the results of clustering image data using k-means clustering. Test data without LBP texture extraction, two alternative results are obtained. The first alternative, cluster 1 members are 24 data and cluster 2 is 6 data. The second alternative, cluster 1 members are 22 data and cluster 2 is 8 data. In the test data with LBP texture extraction, three alternative results are obtained. The first alternative, cluster 1 members are 23 data and cluster 2 is 7 data. The second alternative, cluster 1 members are 17 data and cluster 2 is 13 data. The third alternative, cluster 1 and cluster 2 members are 15 data each.

Item Type:	Article
Subjects:	Q Science > Q Science (General)
Faculty / Institution:	Fakultas Teknologi Informasi
Depositing User:	Fakultas Ekonomi
Date Deposited:	23 Apr 2025 06:16
Last Modified:	23 Apr 2025 06:16
URI:	https://eprints.unisbank.ac.id/id/eprint/10240

Actions (login required)

View Item