Fingerprint Enhancement by Spectral Analysis Techniques

• Authors: Teddy Ko
• Venue: AIPR '02 Proceedings of the 31st Applied Image Pattern Recognition Workshop on From Color to Hyperspectral: Advancements in Spectral Imagery Exploitation
• Year: 2002

This paper presents techniques from spectral analysis thatcan be used for the enhancement and restoration ofdeteriorated latent fingerprints, which do not haveadequate quality for their positive identification. The useof fingerprints as evidence of crime has been one of themost important utilities in forensics since the late 19thcentury. Following the acquisition of a latent fingerprint,there are two operations to be performed: latentfingerprint Enhancement and Matching. To facilitatemore accurate extraction of prominent features in thefingerprint, enhancement has to be performed in order toeliminate the noise and variety of background patterns.Matching is performed by comparing the enhanced latentfingerprint with one of the fingerprints already indatabase. The latent fingerprints are often blurred,incomplete, degraded, and their spatial definition is notclear. These features make their classification andcomparison very difficult if not impossible. For thesolution of this problem, we analyzed some spatial non-linearfilters and frequency domain filters using adaptiveFast Fourier Transform. As a result of the analysis wefound that applying selective filters to the Fourierspectral, the resulting image showed prominent featuresin fingerprints that could not be extracted by otherenhancement techniques.