The Research On Touchless Fingerprint Matching Algorithm

With the rapid development of economy, the information security has shown unprecedented importance in finance, national security, law and politics, e-commerce and e-government application. Identification and authentication are necessary prerequisite to ensure information security. Biometric identification shows great superiority and reliability, due to the utilization of human physiological and behavioral characteristics. And with the super characteristics of uniqueness and invariance, fingerprint identification becomes most popular among other biometric identification technology and has been widely applied and developed.The automated fingerprint identification system (AFIS) technology matures day by day with the development of computer and network technology. However, almost all of the AFIS use touch-based, such as optical and capacitive pressure sensitive, acquisition method. This acquisition method has inherent problems: the fingers must be in close contact with the sensor panel; the irreparable deformation causes by the inconsistent finger pressure; the latent fingerprint on the sensor panel.To fix these problems, we improve the system fundamentally by using the touchless acquisition method. In order to compensate the small area of fingerprint captured by this method, we use three cameras to form a multi-channel acquisition device. For the multi-channel and touch-less fingerprints, we also develop the corresponding feature extraction algorithm, matching algorithm based on local descriptors and fusion algorithm.The experimental result shows that the multi-channel touchless acquisition device and the corresponding recognition algorithm can achieve the desired purpose. Compared to the traditional identification system, our system improves the accuracy and reliability of identification with increasing the time cost. This novel system has good prospects and practical value in access control and other security aspects.