Table 1.

Examples of the methodologies used to process imagery from persistence studies which employ a fluorescent proxy for trace evidence

Year	Paper Title	Purpose of the UV	Image Processing and Analysis Method (Verbatim)
2006	The transfer and persistence of trace particulates: experimental studies using clothing fabrics	To emulate lighter flint particles 1	“photographs were pixelated using Corel Photo‐Paint 9 and the number of particles in each image were computed (as a function of pixel brightness)” (Bull et al. 1: 191)
2009	The Forensic Analysis of Sediments Recovered from Footwear	To explore the movement of silt‐sized particles 4	“This digital image was then pixelated in IDRISI to provide an indication of the amount of silt‐sized material remaining on the sole.” (Morgan et al. 4: 9)
2012	Multiple transfers of particulates and their dissemination within contact networks	To act as a proxy for particulate trace evidence while investigating multiple transfers 3	“The presence of UV powder on the stub was […] quantified using an image rasterisation technique in MATLAB which was specifically adapted for the specifications of this study from Bull et al.[2006]” (French et al. 3: 34)
2013	The recovery of pollen evidence from documents and its forensic implications	As a proxy for pollen on the surface of documents 5	“The digital images taken of each experiment were imported into Coral Photo Paint 11 [sic]. The images were graphically enhanced, pixelated and then five sections of 32 × 7 pixels were counted” (Morgan et al. 5: 377)
2017	Tracers as invisible evidence—The transfer and persistence of flock fibres during a car exchange	UV flock fibres as a tracer 7	“A Matlab (version R2012b) algorithm was created to enable fast automated counting of flock fibres on pictures. The individual pictures were loaded into Matlab and processed automatically. Firstly, the original RGB (red, green, and blue) colour images were converted to a grey value image by extracting the green channel. Subsequently, the foreground (i.e., the flock fibres) was separated from the background (i.e., the target materials) by thresholding. A region of interest (ROI) was selected as well. Next, the fibres were counted. As a result of the varying illumination conditions, the size of one fibre (in pixels) was not the same on all pictures and therefore had to be estimated for each image first. Subsequently, adding up all the foreground pixels and dividing them by the estimated number of pixels per fibre, yielded the amount of fibres on an image.” (Slot et al. 7: 181)