Purchase this article with an account.
Parker Banks, Ralph Gutoskie, Allison Sekuler, Patrick Bennett; Perceptual learning in police fingerprint detectives.. Journal of Vision 2018;18(10):1070. doi: 10.1167/18.10.1070.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
The identification of fingerprints collected at a crime-scene is not a computer automated process. Instead, human experts visually match fingerprints to potential suspects. Although it is known that these fingerprint experts are extremely accurate at visual identification (Thompson, Tangen, & McCarthy, 2011), there has been little research on the perceptual learning that accompanies fingerprint training, and thus little opportunity to increase the speed and efficiency with which experts can be trained and complete their work. Therefore, we tested the spatial contrast sensitivity of police fingerprint experts, as they engaged in a 9-week forensic training program at the Ontario Police College. We found that trained experts become more sensitive to coarse details contained at low spatial frequencies, while they remain relatively insensitive to high spatial frequencies. We conducted additional experiments on the relationship between the information present within fingerprints and their identifiability. We found that it is this low-frequency information that is important in determining print identity. The finding that coarse details are vital to fingerprint indentity contrasts with the common notion that experts rely on specific minutiae (Galton, 1893) to identity prints. To allow for the future study of fingerprint identification, we also describe a new latent fingerprint dataset. This dataset consists of hundreds of fingerprints collected from simulated crime scenes, matched to reference, inked fingerprints taken from the same individual. This dataset is freely available, in order to encourage future study of fingerprint expertise.
Meeting abstract presented at VSS 2018
This PDF is available to Subscribers Only