AFTE Store - Validation Tests for the Congruent Matching Cells (CMC) Method Using Cartridge Cases Fired with Consecutively Manufactured Pistol Slides

The National Institute of Standards and Technology (NIST) is developing the NIST Ballistics Identification System (NBIS) based on three-dimensional (3D) topography measurements and correlations of the paired correlation cells within the correlated topographies[1] to establish ballistics identifications. The Congruent Matching Cells (CMC) method [2] uses three criteria to determine a pair of matching cells. These consist of a threshold for the cross correlation function maximum CCF , and a consistent pattern for registration angle ? and x-y registration position among matching pairs of cells. The number of congruent matching cell pairs C is determined by the three identification parameters associated with their thresholds [2]. Presently the proposed numerical identification criterion for identification is C = 6. That is, two surfaces are identified as being fired by the same firearm if they contain at least six matching cell pairs [1, 2]. In order to test the CMC method and to verify the proposed numerical identification criterion, 40 cartridge cases fired from pistols with 10 consecutively manufactured slides are measured by a 3D confocal microscope and correlated by the CMC method. The breech face topographies are divided into cell arrays (7 X 7 or 6 X 6) for correlation. There are a total of 780 correlations comprising 63 known-matching (KM) and 717 known-non-matching (KNM). Initial tests support both the proposed CMC method and the numerical identification criterion C = 6 for ballistics identifications. Test results show a significant separation between the KM and KNM distributions without any false positive or false negative identification. This represents the highest objective identification accuracy for the same set of cartridge cases that have been tested at NIST thus far. The identification accuracy can be further improved by optimization of the number of cells and the thresholds of the identification parameters, and registration algorithm.