Variability of Bodily Measures of Normally Dressed People Using PhotoModeler Pro 5

• NCJ Number: 225605
• Journal: Journal of Forensic Sciences Volume: 53 Issue: 6 Dated: November 2008 Pages: 1393-1399
• Author(s): Peter K. Larsen M.Sc.; Lone Hansen Ph.D.; Erik B. Simonsen Ph.D.; Niels Lynnerup Ph.D.
• Date Published: November 2008
• Length: 7 pages
• Publication Format: Article
• Publication Type: Report (Study/Research)
• Language: English
• Country: United States of America

Annotation

Since little is known about the accuracy of locating hidden body-points (e.g., the shoulder joint hidden under several layers of clothing) in images obtained with low-resolution surveillance cameras, this study quantified the interobserver and intraobserver variability of bodily measures of clothed individuals in two different poses, determining whether body segment lengths could be used to distinguish between people of similar stature based on an examination of surveillance videos.

Abstract

The study found that points located on the surface of the body (chin, eye, and the outer upper point of the shoulder blade) were the best-reproduced points; the reproducibility was equally good in the intra-observer and inter-observer study. When the points were placed in the joints hidden by clothes, the reproducibility generally decreased, especially in the inter-observer study. The highest variability was found for the points at the hip joint and the straight knee joint in pose 1, where the joint position was difficult to locate because of the loose-fitting trousers in this pose. The wrist joint was recognizable in both poses, and the elbow joint was slightly flexed in all the subjects in pose 1, so the extra flexion in pose 2 did not enhance the process of locating these two joints. Stature and shoulder height could be reproduced within a few centimeters in ideal situations; however, other body segment lengths might not contribute to identification of subjects with similar stature due to the low reproducibility of body-points compared to the expected variability of body segment lengths within subjects of similar stature. In order to improve reproducibility, images of the perpetrator (seen on the surveillance tape) and the suspect (person suspected of being the perpetrator) in the same pose and with flexed joints of the extremities should be used if possible. Precautions should be taken against inter-observer variability. 6 tables, 5 figures, and 19 references