Sign In

Communications of the ACM

ACM TechNews

Scars, Marks, and Tattoos: A Soft Biometric For Identifying Suspects and Victims

View as: Print Mobile App Share:
skull tattoo


A variety of biometric systems capable of matching fingerprints, faces, and irises are already in use by law enforcement agencies to identify suspects and victims. However, there are numerous situations in which primary biometric traits are either unavailable, are difficult to capture, or where the quality of the image is too poor. In such situations, "soft" biometric traits such as height, sex, eye color, ethnicity, scars, marks, and tattoos can help identify a person. Tattoo patterns are regularly cataloged when booking suspects. Based on an ANSI/NIST-ITL (Information Technology Laboratory) standard, each image is manually labeled into one of 70 categories and stored with a suspect's criminal history record. Unfortunately, matching tattoos is a time-consuming, subjective process, and the simple class descriptions do not include all the semantic information present in an image. SPIE has developed an automatic tattoo matching system called Tattoo-ID, which uses content-based image retrieval based on tattoo features, such as color, shape, and texture, instead of labels or keywords.

Tattoo-ID retrieval experiments In these two examples of retrieval experiments, each row contains a query image followed by the most-similar images that Tattoo-ID found in database. The correct retrieved image(s) for each query are enclosed in red square boxes. Credit: SPIE 
Tattoo-ID provides users with a group of images that most closely resemble the queried tattoo. User feedback, based on the retrieved images, can be used to refine feature extraction and the matching capabilities. Tattoo-ID also uses class and subclass labels so users can specify the tattoo image and ANSI/NIST categories, to keep the system compatible with current law enforcement practices.

View Full Article


Abstracts Copyright © 2009 Information Inc., Bethesda, Maryland, USA


No entries found

Sign In for Full Access
» Forgot Password? » Create an ACM Web Account