Firing Pin Impression Segmentation using Canny Edge Detection Operator and Hough Transform

Abstract

Firearms identification based on the forensic ballistics specimen is crucial in solving criminal case in a short time. Currently, the firearms examiners perform authentication by visual observation. Due to observation of large evidence database, the experts normally take a long time to identify the firearms. As a result, computerized firearms identification should be implemented in order to perform the identification faster. The computerized identification involves image preprocessing, segmentation, feature extraction and classification. Therefore, in order to reduce computational time, the segmentation has to be performed automatically. The main objective of this study is to perform the segmentation of firing pin impression by using Canny edge detection operator improvised with Hough transform. The performance of segmentation in detecting the central image of firing pin impression has achieved 93% segmentation accuracy