Mathematical Morphology Approach for Genuine Fingerprint Feature Extraction.

International Journal of Computer Science and Security

Ramesh Manza¹  K. V. Kale¹  Vikas Humbe¹  S. S. Gornale¹ 

[1] $$

关键词: Feature Extraction;    Morphology;    Fingerprints;    superfluous information etc;   

DOI  :  

来源: Computer Science and Security

【 摘 要 】

Recognition systems based on biometrics (faces, hand shapes and fingerprintsetc.) are finally taking off although it has taken a long way to come. Fingerprintshave been a precious tool for law enforcement, forensics and recently incommercial use for over a century. Evaluate the performance of these emergingtechnologies is tricky problem. Most fingerprint verification algorithms rely onminutiae features, and these algorithms can only be as robust as the underlyingminutiae features. Therefore, reliable minutiae extraction is vital to a system’sperformance. Most of the feature extraction techniques extract features fromthinned images but while dealing with binarization and skeletization of image itintroduces noise or superfluous information, which creates troubles for genuinefeature extraction. In this paper we have used the mathematical morphology toremove the superfluous information for genuine feature extraction and measurethe feature extraction performance through sensitivity and specificity.

【 授权许可】

Unknown