A Hybrid Approach to Validate and Retrieve Fingerprint Image for Authenticating Database Sensitive Information Access and to Verify Some of Topological Multi-Granular Rough Sets Properties on Authenticated Database Fields

Content based image retrieval has already begun to turn into an intriguing and exceptionally needed field to meet with the client prerequisite. Fingerprint approvals likewise have become a fascinating field to principally uphold the confirmation framework. Notwithstanding foundations regulating guaranteed about information straightforwardly, affiliations are also obliging them to examine the fingerprint endorsement. Despite establishments overseeing ensured about data, directly associations are similarly obliging them to research the fingerprint check. Organizations and foundations both are typically utilizing the idea of fingerprint for the participation and presence reason. Nevertheless, establishments presently managing secured information as like institutes and additionally going along with them to investigate the fingerprint authentication. More than this usual action there are some particular applications where the authentication would be considerable to help the application than using a traditional pattern of password. We proposed a fingerprint authentication system for the scientific based industry database access system. We removed the noise if any found in the input fingerprint using anisotropic Gauss filtering. The noise free fingerprint image is not further passed to the binarization phase as the need of binarization and some form of the drawback of binarization is replaced with a thinning algorithm. The thinned fingerprint image is followed by the matching process and once it is matched, then only the sensitive and highly confidential information are allowed to access from the database. The testing is done using OpenCV and FVC 2004 dataset to assess the projected retrieval method. Further the matched fingerprints are used to prove some of the internal structure properties of multi-granular rough sets (MGRS) on authenticated database fields. The accuracy obtained in this hybrid approach is found to be more satisfactory and hence the security towards accessing the sensitive information is well achieved.