Conceptual design of firearm identification mobile application (FIMA)

摘要

Everybody has at least a smartphone and the technology keep on evolving from time to time, same goes to mobile applications. Mobile applications are currently more towards learning, business and entertainment purposes, but why not we extend the use of the technology to forensic investigation? In this paper, we are highlighting the potential of a mobile application on smartphones to support investigation in crime scenes. As we know, crimes involving firearms has been a threat to humanity since time immemorial. The mass production of handguns and other firearms since the nineteenth century has similarly expanded crime numbers globally. Thus, it is critical to identify the firearm used whenever a firearm crime has been reported, which calls for expertise in forensic ballistics. Firearms identity is a crucial goal of firearm analysis. From previous research, the firing pin impact impression on a cartridge case from a fired bullet is one of the most extensive clues in firearms identification. Since nowadays a lot of crime cases involve guns, a portable device should help the police and the forensic teams for immediate firearm identification, whereby they can pre-analyze the firearm impression on the spot to gain some prior information, which is cost, space and time saving. In this paper, mobile application for firearm identification is proposed, named Firearm Identification Mobile Application (FIMA). This firearm identification application will analyze numerical features extracted from the firing pin impression image as the unique features, and immediately catalogues the features using the backpropagation neural network classification technique. It is found that 'trainlm' performed the best compared to the other training algorithms based on average overall correct classification rates for all features. On the other hand, 'trainscg' performed the best based on centre firing pin impression images. In the near future, a portable firearm analysis device will be developed, namely Portable Firearm Analysis Device (PAFAD) to further assist in forensic investigation activities.