Wet and Wrinkled Finger Recognition using Voting Classification and GLCM Algorithm

by Puneet Kumar, Saurabh Sharma (Foreword)

The automated wrinkle fingerprint recognition techniques rely on some principles from the domain of pattern recognition. Forward many schemes for minutiae extraction. However, these schemes perform well only with clean and good quality images. Wrinkle refers to a pattern in the fingerprint that can cause fake minutiae. Wrinkles are a type of stripes that cross ridges and valleys erratically in fingerprints. They occur due to aging, manual labour, accidents and so on. Some wrinkles are everlasting, while others are temporary, and disappear after some time. The wrinkles of both types present counterfeit minutiae and therefore, the performance of the fingerprint recognition system was degraded. Two prevalent techniques are presented for recognizing the wrinkle fingerprint. The first method is planned on the basis of pixel intensity of the wrinkles, and the second relies upon the pattern of fingerprint orientation fields. In this research paper classification method is proposed for the wet and wrinkled image recognition. The proposed method is the combination of threshold segmentation, Grey Level C0-occurrence Matrix algorithm to extract the attributes and voting classification to recognize wet image. The proposed method is executed in python and the analysis of results is done with regard to accuracy, precision and recall. The accuracy of the proposed model is achieved up to 96 percent.

91 pages, Kindle Edition

Published February 10, 2022