NCJ Number
100954
Date Published
1985
Length
12 pages
Annotation
This British evaluation of an automatic system of signature authentication found that it has numerous features which make it unsuitable for forensic applications in any of its present forms.
Abstract
The device scans the signature of persons opening checking accounts and records a number of parameters. These parameters are reduced to a series of digits printed on the issued checks. The device assesses the compliance of the signatures on issued checks with the checks' printed digits. Thresholds for acceptance or rejection can be set by the bank to provide an acceptable level of forgery detection. To test the device in a forensic situation, several batches of signatures were analyzed. Some genuine signatures were rejected; some freehand copies were accepted; most traced copies were accepted; most disguised signatures were rejected; and in case-related applications, conflicting results were obtained. The device's weakness is its limited focus on the signature's pictorial effect. Document examiners, on the other hand, focus on letter design, line quality, stroke sequence, pressure variation, and other factors not assessed by the device. 4 references, tabular data, and specimens.