Абстрактный

The improved fingerprint enhancement algorithm research based on embedded environment

Feng Wang, Bo Han, Tian Peng Han

For the problem that existing fingerprint enhancement algorithms computational complexity which is inappropriate in embedded environment, the author proposes polyline large template enhancement algorithm. The algorithm has the following steps: Firstly, we decompose the two-dimensional Gabor function in vertical direction, and combine Bresenham linear fast generate algorithm to realize a one-dimensional Gabor fast enhancement. Secondly, we improve one dimensional fast Gabor template by using a polyline large template method. The algorithm not only improve the enhancement effect in the serious curvature area, but also get a better connection for broken line. Experiments use FVC2004 fingerprint database to validate, the experimental results demonstrate the effectiveness of the algorithm