Given a database of fingerprint images, our task is to reduce the rejection rate by enhancing the images.
Fingerprints taken by line sensors are inherently of bad quality and biased with certain kind of noise. In particular partly destroyed regions of the skin and folds are reducing the detection performance considerably.
In this project we are concerned with the following tasks:
- Improving the general image quality
- Seperation of ridges and folds and removal of folds by imprinting
- Improving other partly recognizable regions
- Considering uncertainty in the fingerprint matching algorithms.