Tong Mingsi, Song John, Chu Wei, Thompson Robert M
School of Mechatronics Engineering, Harbin Institute of Technology, Harbin 150001, China ; National Institute of Standards and Technology, Gaithersburg, MD 20899, USA.
National Institute of Standards and Technology, Gaithersburg, MD 20899, USA.
J Res Natl Inst Stand Technol. 2014 Nov 6;119:575-82. doi: 10.6028/jres.119.023. eCollection 2014.
The Congruent Matching Cells (CMC) method for ballistics identification was invented at the National Institute of Standards and Technology (NIST). The CMC method is based on the correlation of pairs of small correlation cells instead of the correlation of entire images. Four identification parameters - T CCF, T θ, T x and T y are proposed for identifying correlated cell pairs originating from the same firearm. The correlation conclusion (matching or non-matching) is determined by whether the number of CMC is ≥ 6. This method has been previously validated using a set of 780 pair-wise 3D topography images. However, most ballistic images stored in current local and national databases are in an optical intensity (grayscale) format. As a result, the reliability of applying the CMC method on optical intensity images is an important issue. In this paper, optical intensity images of breech face impressions captured on the same set of 40 cartridge cases are correlated and analyzed for the validation test of CMC method using optical images. This includes correlations of 63 pairs of matching images and 717 pairs of non-matching images under top ring lighting. Tests of the method do not produce any false identification (false positive) or false exclusion (false negative) results, which support the CMC method and the proposed identification criterion, C = 6, for firearm breech face identifications using optical intensity images.
弹道识别的全等匹配单元(CMC)方法是由美国国家标准与技术研究院(NIST)发明的。CMC方法基于小相关单元对之间的相关性,而非整个图像的相关性。提出了四个识别参数——T CCF、Tθ、Tx和Ty,用于识别源自同一枪支的相关单元对。相关性结论(匹配或不匹配)由CMC的数量是否≥6来确定。该方法此前已使用一组780对三维地形图图像进行了验证。然而,当前本地和国家数据库中存储的大多数弹道图像都是光强度(灰度)格式。因此,将CMC方法应用于光强度图像的可靠性是一个重要问题。在本文中,对在同一组40个弹壳上采集的枪膛表面印记的光强度图像进行相关性分析,以使用光学图像对CMC方法进行验证测试。这包括在顶环照明下63对匹配图像和717对不匹配图像的相关性分析。该方法的测试未产生任何错误识别(误报)或错误排除(漏报)结果,这支持了CMC方法以及所提出的用于使用光强度图像进行枪支枪膛表面识别的识别标准C = 6。