Font Size: a A A

Research On Personal Identification Based On The Fusion Of Multi-finger Knuckleprints

Posted on:2011-07-13Degree:MasterType:Thesis
Country:ChinaCandidate:R ZhaoFull Text:PDF
GTID:2178360308954379Subject:Communication and Information System
Abstract/Summary:PDF Full Text Request
Biometrics is a automatic personal identification technology which uses the unique physical or behavioral characteristics of a person. Because the commonly used biometrics have defects themselves, the study of new biometric is of great value. A relative new biometric—knuckleprint is choosed to study. Unlike most previous work, there is no need to collect a large number of images to train classifirer in the experiment, the model database is composed of only single image from every-person. We complete the experiment of knuckleprint's line feature recognition and multiple finger-knuckleprints decision level fusion. The results of the experiment show that the method proposed in this paper is effective.The research work are as follows:1. According to the characteristic of the knuckleprint: stable, one direction and only few simple main lines, we proposed a new gradient operator to extract the line feature of the knuckleprint from the low-resolution finger images.2. When the test image and the template are matched, the dislocation often occurs between the two images. In order to reduce the impact on the identity decision of the dislocation, we proposed a new matching-method. By translating the test image 8 times in the neighborhood of one pixel, then we use these 9 images matching with the template, then take the biggest matching-score as the basis of the decision for the test image identity.3. Based on the decision fusion technology, we proposed a multimodel biometric using four fingers'knuckleprints from a person to determine the identity of the user. The experiments show that the correct recognition rate of the system is well improved.
Keywords/Search Tags:multi-fingers, knucklerprint, line features, base on single image, PCA, decision based fusion
PDF Full Text Request
Related items