Quantitative Evaluation of Footwear Evidence: Advancing the Footwear Impression Comparison System (FICS) towards Casework Application

Analysis of footwear evidence is an active area of research. Several groups, including Dalian and CSAFE, have produced new approaches for automatic make/model identification. Amidst this research activity, we believe NIST’s efforts are leading the way towards aiding examiners in casework evidence evaluation. NIST was funded by NIJ in 2017 to conduct research on quantitative methods for 2-D footwear evidence assessment that could assist examiners by providing empirical support for their findings. We developed a prototype system called the Footwear Impression Comparison System (FICS) that we believe is nearly ready for broad testing on case-like evaluations. FICS parallels the workflow used by examiners, producing a sequence of evaluations of size and design correspondence, wear correspondence, and RAC correspondence, taking into consideration the clarity and completeness of the questioned impression. The algorithm results can be mapped to corresponding conclusions following, for instance, the SWGTREAD conclusion scale. The prototype system has been successfully applied to comparisons involving full and partial crime scene impressions made on a variety of substrates, matrices, and collection methods seen and used in casework. These include, among others, chemically enhanced synthetic blood on tile, an electrostatic lift from paper, and a gel lift from dusty paper. While preliminary evaluations have shown promising potential for each of the prototype components, further testing is required to demonstrate performance and identify limitations across the wide range of crime scene impressions. This additional testing will lead to further algorithm improvements and prepare FICS for the ultimate goal of this proposal, which is to produce an easy-to-use interface and participate alongside examiners in the FBI footwear black box study. If successful, the research described in this proposal will lay the groundwork for the first application of quantitative algorithms to footwear impression evidence evaluation in actual casework. The major project tasks are (1) collect and analyze impressions from at least 200 staged crime scenes (and corresponding test impressions) for algorithm testing, (2) automate clarity markups of crime scene impressions and refine other workflow components to address limitations identified during testing, (3) use FICS to participate in the FBI black box study. These tasks will be implemented over a two-year time frame. NIST is partnering with footwear examiners from the FBI and will be working closely with CSAFE. Expected products include conference presentations, scientific papers, seminars, training workshops, tutorial videos, and implementations of all our algorithms that could be used by others.