the Forensic Test Information System Firearms Ammunition Protective Clustering Process Using Gray Level Co-occurance Matrix and K-Mean Clustering Algorithms

didik supriyadi

JURNAL TEKNIK INFORMATIKA DAN SISTEM INFORMASI
Vol 13 No 2 (2026): JATISI (Jurnal Teknik Informatika dan Sistem Informasi)

The use of technology is a solution when developments continue to increase and develop. The connection between technology and the field of state security is no exception. The method that supports clusterization is feature extraction using the Gray Level Co-occurrence Matrix (GLCM), carried out before the clusterization process. GLCM is suitable for extracting features or characteristics in images with unique patterns, such as puppet pattern recognition research. This research procedure is a flow chart to build an information system for forensic testing of the clustering process for firearm ammunition projectiles using the Gray Level Co algorithm. -occurrence Matrices (GLCM) and K-means clustering. Figure 3.1 below illustrates the information system framework as an explanation of each input flow, process, and output. The research results show that the GLCM method for feature extraction from grayscale images and the K-Means method for clustering provide good results and accuracy. The model performance reached 71.14% even with limited data. This model can be used in console applications such as Google Collabs and with a GUI with relatively stable application performance.