ANALISIS KINERJA METODE GLCM DAN LS-SVM DALAM KLASIFIKASI CITRA SAMPAH ORGANIK DAN ANORGANIK

Michelyn Angela Sabatini Rajagukguk (Universitas Bina Sarana Informatika)
Ahmad Fauzi (Universitas Bina Sarana Informatika)
Bambang Wijonarko (Universitas Bina Sarana Informatika)

Scientific Journal of Informatics
Vol. 10 No. 2 (2025): Jurnal Ilmiah Informatika

Waste management, particularly in distinguishing organic and inorganic types, remains a major environmental challenge. Manual sorting processes are inefficient and prone to errors. This study aims to develop an automated waste image classification system using a combination of Gray Level Co-occurrence Matrix (GLCM) and Least Squares Support Vector Machine (LS-SVM). A total of 1,060 images were used, divided equally between organic and inorganic categories. Texture features such as contrast, correlation, energy, and homogeneity were extracted using GLCM and combined with mean RGB color features. The LS-SVM model with the Radial Basis Function (RBF) kernel achieved an accuracy of 87 percent, outperforming conventional SVM. The model’s effectiveness aligns with previous studies that used SVM-based waste classification and texture feature enhancement with GLCM descriptors. The model was implemented using a Flask web application for real-time predictions.