Bullet identification based on striation features using fast fourier transform and artificial neural network
Loading...
Date
2017-06
Authors
Chan, Zhen Yu
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Firearms identification from bullet specimens is important and useful in crime and forensic investigation. When a bullet is fired, characteristic markings are created due to contact between the bullet and the barrel of the gun. Every firearm has its own unique characteristic markings, also called ‘fingerprint’ regardless of its size, type and model. These unique characteristics are the important features in identifying firearms. However, traditional bullet identification is a labor intensive activity with several weeks of time being devoted to a single analysis and comparison. This paper investigates the firearm identification method based on fast Fourier transform (FFT) and Artificial Neural Network (ANN). This project presents an approach to examine and analyze the bullet specimens using FFT technique for 9 mm handgun identification. There are 5 types of bullets classes and 6 specimens per class. A total of 30 specimens bullet were used in the identification method and they were scanned by using Alicona Infinite Focus measurement machine. Fundamental frequency and harmonics were extracted by the FFT technique and act as input parameter for neural network training. ANN was applied in this project to identify the bullet classes. Experimental result show that the proposed system can achieve 66.7% accuracy through analyzing the fundamental frequency and harmonics of the fired bullet specimens. Although the amplitude of the fundamental frequency and harmonics have limited impact to justify the bullet type but there is still room for improvement of the classification accuracy.