Font Size: a A A

Fingerprint Identification System Algorithm

Posted on:2007-06-08Degree:MasterType:Thesis
Country:ChinaCandidate:C W LinFull Text:PDF
GTID:2208360182992515Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
With development of society, the rapid progress of computer science, an accurate and reliable personal identification technology becomes more and more important. While so many drawbacks in traditional methods, people turn to biometrics, a new research field with bright future. Fingerprint identification, one of the most important branches in biometrics, is showing its unique advantages. Nowadays, automatic fingerprint identification technology has been applied to many domains widely such as police, custom, bank and network security etc. Thus, to execute deep researches on fingerprint recognition is not only theoretically meaning but also practically meaning.Using technology of image processing and pattern recognition, the thesis discusses and studies a number of problems about automatic fingerprint identification system. The thesis achieves fingerprint pretreatment, minutiae extraction and fingerprint matching etc and brings forth new ideas in fingerprint division, fingerprint enhancement. The main contents are as follows:1. In the process of fingerprint pretreatment, we present a new fingerprint division algorithm based on variance and mathematical morphology. This algorithm is not only very effective, but also simple in computation. Then, we adopt that we present an algorithm based on region for fingerprint direction image postprocessing in order to get an accurate fingerprint direction image. Base on the accurate fingerprint direction image, we can get a satisfactory effect of fingerprint enhancement. Next, the gray image is transformed into a binary image through a section-threshold binary algorithm. Finally, thinning is implemented on the given binary image.2. Algorithms for feature extraction of fingerprint are studied. Minutiae can be extracted with exactitude by 8-neighborhood cross-calculation algorithm, which is a kind of algorithms has good ability of anti-jamming. Then, a feature extraction postprocessing algorithm based on ridge tracing is adopted. It can remove pseudo-structures like spurs, bridges and holes etc from a thinned fingerprint image. After this step, spurious minutiae is not a trouble to us anymore. So far, a...
Keywords/Search Tags:fingerprint identification, fingerprint division, fingerprint enhancement, feature extraction, fingerprint matching
PDF Full Text Request
Related items