NCJ Number
240265
Journal
Journal of Forensic Sciences Volume: 57 Issue: 3 Dated: May 2012 Pages: 610-617
Date Published
May 2012
Length
8 pages
Annotation
This study examined the estimation of bloodstain volume using fractal analysis of digital images of passive blood stains.
Abstract
The estimation of bloodstain volume using fractal analysis of digital images of passive blood stains is presented. Binary digital photos of bloodstains of known volumes (ranging from 1 to 7 mL), dispersed in a defined area, were subjected to image analysis using FracLac V. 2.0 for ImageJ. The box-counting method was used to generate a fractal dimension for each trial. A positive correlation between the generated fractal number and the volume of blood was found (R2 = 0.99). Regression equations were produced to estimate the volume of blood in blind trials. An error rate ranging from 78 percent for 1 mL to 7 percent for 6 mL demonstrated that as the volume increases so does the accuracy of the volume estimation. This method used in the preliminary study proved that bloodstain patterns may be deconstructed into mathematical parameters, thus removing the subjective element inherent in other methods of volume estimation. Abstract published by arrangement with John Wiley & Sons.