Assessing similarities firing pin traces of firearms on their digital images

The mathematical method for comparing images of firing pin traces marked by arbitrary shape spots in grey scale representation is considered. The algorithm based on consecutive application of potential function and contour analysis methods is proposed. The comparison algorithm final results do not depend on image orientation. The ability to predict the possible variability of individual features in digital images firing pin traces by their binarization at different levels and fixing changes of the Euler's characteristics is shown. The effectivity of the proposed algorithm has been proved through an array of 120 digital images of similar traces.
The offered method can be used as an independent for carrying out checks on an array of digital images of firing pin traces, and as a preliminary for one of correlation methods. The work is of interest to software developers of automated ballistic identification systems.