Overlapped fingerprint image capture and separation using digital holography and machine learning

Even now that DNA test has become common, latent fingerprints are still a useful evidence for finding criminals. However, the latent fingerprints are often damaged and overlapped with each other. At present, these overlapped fingerprints are separated manually by a person. It is not only time-consuming task, but also less likely to be adopted as evidence because it is difficult to rule out of the intention of a person. In this paper, non-destructive method to capture and separate overlapped fingerprints using digital holography and machine learning is presented and demonstrated.