Contactless automated lifting of latent fingerprints from difficult curved surfaces

Abstract

We present a fully automated method for contactlessly extracting latent fingerprints and correcting for surface curvature distortions. Our proposed method is rapid, automatic, zero-contact and chemical-free, and is able to obtain high quality fingerprint images from (forensically) difficult curved surfaces that are specular and/or transparent, such as glass lightbulbs and chrome water taps. Its contactless nature preserves the prints for later DNA analysis, and we demonstrate that the automatic surface curvature distortion correction improves fingerprint matching scores, often significantly, across a range of objects commonly found at a crime scene.