期刊名称：International Journal of Innovative Research in Science, Engineering and Technology
印刷版ISSN：2347-6710
电子版ISSN：2319-8753
出版年度：2016
卷号：5
期号：6
页码：11553
DOI：10.15680/IJIRSET.2015.0506267
出版社：S&S Publications
摘要：Fingerprint recognition is the most widely used biometric to identify individuals. Latent fingerprintidentification is of critical importance in criminal investigation. The system is focused on the dealing the issues ofpoor quality with unclear ridge structure and various overlapping patterns for the latent fingerprint images. Hence themethods are such as total variation (TV) model and multiscale patch-based sparse representation introduced toimprove the reliable feature extraction and recognition as well as improve the image quality. It is also used toimprove the sparse representation in image denoising. The specified image is decomposed into cartoon and texturecomponent then apply the TV model on the cartoon component to remove the structured noise as well as multiscalepatch-based sparse representation technique for the enhancement of the texture component. The Gabor dictionariesare constructed to capture the uniqueness of fingerprint ridge structure. Also multiscale patch-based sparserepresentation is iteratively applied to reconstruct high-quality fingerprint image. However this system is failed toachieve the global ridge structure for low quality latent fingerprints and hence reliability is reduced as well as systemperformance is degraded. To achieve the global ridge structure quality, an optimization algorithm named as fireflyalgorithm. Fingerprint image enhancement is an essential preprocessing step to extract qualitative minutiae from afingerprint image. Image enhancement is mainly done by maximizing the information content of the enhanced imagewith intensity transformation function. The firefly algorithm is used to update the best features globally and globaloptimization is increased. This system is greatly reduced the noise rates in the specified latent fingerprint images andimproves the global ridge structure significantly. From the experimental result, the conclusion decides that the globaloptimization provides higher performance rather than total variation (TV) model and multiscale patch-based sparserepresentation in terms of high image quality.
