NCJ Number
229279
Journal
Journal of Forensic Sciences Volume: 54 Issue: 6 Dated: November 2009 Pages: 1365-1375
Date Published
November 2009
Length
11 pages
Annotation
In order to test whether the difference in the measurement of a person's height in an image and that person's actual height is due to the camera and other circumstances, an experiment was conducted that involved the measurement of 22 test persons using 3 cameras of varying quality.
Abstract
The main conclusion of the study is that when height measurements are performed on images of persons, an error is made that depends on the stance of the person and on the camera position. A statistical analysis of measurement on different persons must be performed for each new case and helps to gain more insight regarding this error. Reproducibility of measurement per image was apparently strongly dependent on the camera quality; whereas, systematic bias differed with the view point of the camera. The measurement process was also dependent on the camera operator, so its repetition by different operators is recommended. Apart from the camera quality, the position of the person in the image, and the operator, other factors are the height of the person in the image in pixels ("resolution" of the person), camera location and orientation (tilt), and lens distortion. As for height measurements in real life, they are always estimates of actual values. Descriptions of methods and techniques address the set-up of the main experiment, notation, kernel density estimation, and numerical analysis. 9 tables, 12 figures, and 9 references