Statistical Validation of the Individuality of Guns Using 3D Images of Bullets

• NCJ Number: 213674
• Author(s): Benjamin Bachrach
• Date Published: March 2006
• Length: 66 pages
• Grant Number(s): 1997-LB-VX-0008
• Agencies: NIJ-Sponsored
• Publication Type: Report (Study/Research)

Annotation

This study conducted by Intelligent Automation, Inc. (IAI) and supported by the National Institute of Justice (NIJ) set out to improve and make advances on the state-of-the-art automated ballistic analysis systems and to develop and validate methodologies for ballistic identification.

Abstract

Overall, this study provides a solid validation of the foundations of ballistic identification. Structured into three main components, the first component of the study, which examined the effect of barrel wear, found that barrel wear did not pose a significant challenge to firearms identification. Component two, dealing with the development of methodologies to address both the evaluation of the degree of individuality of barrels and the estimation of probability of error in bullet-to-barrel classifications found that it was feasible to apply a classification approach to identify bullets fired by different barrels of the same manufacture and make. The third and final component of the study, which examined the degree to which the conclusions of the prior two sections could be applied to damaged bullets, found that it was possible to link a damaged bullet to a barrel with a high degree of certainty. However, the achievable overall classification accuracy was inferior to that attained with pristine bullets. Supported through a grant under the U.S. Department of Justice, National Institute of Justice (NIJ), the Intelligent Automation, Inc. (IAI) conducted a study to improve on the state-of-the-art automated ballistic analysis systems in 3D-based comparison methodologies and take these improvements and make them available to the law enforcement community. In addition, it attempted to develop and make use of a 3D-based ballistic analysis system in order to provide answers to several questions: (1) what qualitative criteria should be used to establish the individuality of a gun; (2) what quantitative criterion should be used to establish that a suspect gun fired a given evidence bullet; and (3) can the probability of a bullet/gun match being erroneous be estimated? Figures and tables