|
Abstract:
|
When a gun is fired, it leaves marks on cartridge cases that are thought to be unique to the gun. In current practice, firearms examiners inspect two cartridge cases using a comparison microscope, and if there is "sufficient agreement," they conclude that they come from the same gun, and testify in courts as such. A recent publication (2016) by the President's Council of Advisors on Science and Technology questioned the scientific validity of such analysis. One recommendation was to convert firearms analysis to an objective method, through the use of fully automated methods. We propose one such method for comparing breech face marks on cartridge cases using 2D optical images. We improve on existing methodology by automating the selection of marks, as well as removing the effects of circular symmetry. The latter is done by fitting a linear combination of basis functions, and this methodology can be more generally applied to any image or situation for which this is desired. Here we find that this reduces the similarity between images that are known to come from different guns. Lastly we propose a method of computing a "random match probability" empirically, given a known database.
|
