Capturing high-resolution digital images for use in forensic document examination.

In the past, pattern disciplines within forensic science have periodically faced criticism due to their subjective and qualitative nature and the perceived absence of research evaluating and supporting the foundations of their practices. Recently, however, forensic scientists and researchers in the field of pattern evidence analysis have developed and published approaches that are more quantitative, objective, and data driven. This effort includes automation, algorithms, and measurement sciences, with the end goal of enabling conclusions to be informed by quantitative models. Before employing these tools, forensic evidence must be digitized in a way that adequately balances high-quality detail and content capture with minimal background noise imparted by the selected technique. While the current work describes the process of optimizing a method to digitize physical documentary evidence for use in semi-automated trash mark examinations, it could be applied to assist other disciplines where the digitization of physical items of evidence is prevalent. For trash mark examinations specifically, it was found that high-resolution photography provided optimal digital versions of evidentiary items when compared to high-resolution scanning.

Accuracy and reliability of forensic handwriting comparisons.

Forensic handwriting examination involves the comparison of writing samples by forensic document examiners (FDEs) to determine whether or not they were written by the same person. Here we report the results of a large-scale study conducted to assess the accuracy and reliability of handwriting comparison conclusions. Eighty-six practicing FDEs each conducted up to 100 handwriting comparisons, resulting in 7,196 conclusions on 180 distinct comparison sets, using a five-level conclusion scale. Erroneous "written by" conclusions (false positives) were reached in 3.1% of the nonmated comparisons, while 1.1% of the mated comparisons yielded erroneous "not written by" conclusions (false negatives). False positive rates were markedly higher for nonmated samples written by twins (8.7%) compared to nontwins (2.5%). Notable associations between training and performance were observed: FDEs with less than 2 y of formal training generally had higher error rates, but they also had higher true positive and true negative rates because they tended to provide more definitive conclusions; FDEs with at least 2 y of formal training were less likely to make definitive conclusions, but those definitive conclusions they made were more likely to be correct (higher positive predictive and negative predictive values). We did not observe any association between writing style (cursive vs. printing) and rates of errors or incorrect conclusions. This report also provides details on the repeatability and reproducibility of conclusions, and reports how conclusions are affected by the quantity of writing and the similarity of content.

An objective inter-comparison of trash mark constellations generated by manual and automated detection methods.

Trash marks are unintentional markings observed on printed, scanned, or photocopied documents that result from permanent defects or transient material in office machines and can be used for source attribution of questioned documents. Trash mark examinations have been in use in forensic laboratories for decades, yet the method remains relatively untested and relies on training, experience, and anecdotal information to support its validity. This study generated and harnessed objective data to empirically test one of the foundational theories for assessing the origin of photocopied documents: provided trash marks are present in sufficient quantity and/or quality, no two machines will exhibit a constellation of trash marks that is indistinguishable from another. In this project, objective trash mark location and size data was generated for 50 known photocopiers using both a traditional and a novel, automated method. Inter-machine comparisons were conducted using a novel variant of the Hausdorff distance algorithm to generate a quantitative assessment of how similar or different the 2450 pairs of trash mark constellations were from one another. This study found that each of the machines bearing one or more trash marks exhibited objective differences in their trash mark constellations, ultimately providing support for the tested hypothesis.