The extent to which garments affect the assessment of body shapes of males from faceless CCTV images.

Abstract

Closed circuit television (CCTV) systems are being widely used in crime surveillance. The images produced are of poor quality often face details are not visible, however expert witnesses in the field of biological anthropology use morphological descriptions of body shapes in an attempt to identify persons of interest. These methods can be applied to individual images when other cues such as gait, are not present. Criminals commonly disguise their faces, but body shape characteristics can be used to distinguish a person of interest from others. Garments may distort the body shape appearance, thus this study was undertaken to investigate the effects of garments on the description of body shape from CCTV images. Twelve adult males representing a wide body shape range of Sheldonian somatotypes were photographed in identical garments comprising of tight fitting black shirt, horizontally striped shirt, padded leather jacket and in naked torso. These photographs were assessed by 51 males and females aged 18-50 years, with varying levels of education, and different experience in use of CCTV images for identification of people, to identify the 12 participants. The effect of assessors was not significant. They correctly distinguished 88.6% of individuals wearing the same wear, but could not match the same individuals wearing different wear above the random expectations. However, they matched somatotypes above random expectation. Type of clothing produced little bias in somatotype matching; ectomorphic component of individuals wearing black shirts and padded jackets was overestimated and underestimated, respectively. In conclusion, type of the wear had little effect in the description of individuals from CCTV images using the body shapes.