Font Size: a A A

Research Of Automate Fingerprint Identification Algorithm

Posted on:2010-03-21Degree:MasterType:Thesis
Country:ChinaCandidate:L L YuanFull Text:PDF
GTID:2178360275484809Subject:Control theory and control engineering
Abstract/Summary:PDF Full Text Request
The fingerprint identification has become one of the most reliable,convenient and popular personal authentication technologies due to its uniqueness an invariability. Paper has done the research as follows: There are some limitations in the fingerprint segmentation based on variance or direction. According to the character that edge map's ridge is difference from background, combing to morphological filter, the paper proposed a new segmentation algorithm based on the character of edge; the enhancement based on sobel masks makes false characters. According to the width of ridge, combining to the character of eight directs, the paper proposed improved masks can enhance the ridge of eight directions by turn; Choice a simply and effective binarization algorithm, based on the difference feature of improved enhancement algorithm; Simplifying minutia extraction. Experiment results show that the system work with celerity,accuracy and reliability.
Keywords/Search Tags:edge detection, morphological filter, fingerprint enhancement, binarization, minutia extraction
PDF Full Text Request
Related items