Shape Adaptation with application to Fingerprint Image Enhancement
Abstract
Fingerprint Image Enhancement can be seen as a scale-space problem. Dense finger print images could sufficiently be analyzed by decomposing each division of image into a set of simpler images at different scales. This purpose can be achieved by the use of appropriate scale space operators. Once image is decomposed, it can be addressed by employing an explicit shape adaptation algorithm as proposed by Almansa and Lindeberg. The whole process has two phases. One has to do with selection of scale for given fingerprint image. the other has to do with adaptation of the local features. However, these two processes need to be optimized jointly. The report deals with the detail of this whole process along with objective results.