PENGENALAN SIDIK JARI MANUSIA DENGAN MATRIKS KOOKURENSI ARAS KEABUAN (GRAY LEVEL CO-OCURRENCE MATRIX)

Abstract

Biometric system is a recognition process by measuring of distinguish traits on human body or human behaviour, then comparing it with one in the database. One of human traits which is unique and specific, stable for long period, and measureable is fingerprint. The purpose of this research is to make application that is able to recognize fingerprints using Gray Level Co-ocurrence Matrix (GLCM) method and observe the recognition level of them. Method used in this fingerprint recognition research is texture analysis. Firstly, fingerprint is captured using fingerprint scanner. Then, the fingerprint image is extracted using Gray Level Co-ocurrnce Matrix (GLCM) featuring. The features obtained are energy, contrast, entropy, maximum probability, homogeneity, dissimilarity, mean, and correlation. The next step is recognition using normalized euclidean distance. Some tests performed in this research are the influence of GLCM’s angle, GLCM’s distance, and number of sample in database. From the previous tests mentioned, the highest recognition, 83,3 percent, is achieved by using all of GLCM’s angle (0°+45°+90°+135°) and 1 pixel distance with twelve samples stored in database.