An Approach for Privacy-Enhancing Actions Using Cryptography for Facial Recognition on Database

Abstract

In the era of digital security, face recognition is highly demanded with other biometric strategies like fingerprints, iris recognition, and hand geometry for identification. In these existing ways, there are some disadvantages like time taken, pose, illumination problem, and also age effects. With this, it has been noticed that the stored images are not secured, and it is possible of cheating with the restriction in non-tolerable areas of security. Keeping these aspects in mind, this work has proposed mixture of two popular methods with cipher images to be saved in the database. Proposed method also focusing on fastest key generation for encryption and can deal with more subjects and less false match rate. Both principal component analysis and local binary pattern algorithms are utilized, and encryption system has been included for storing the images and for recognition in this proposed method. Reducing the measurement of image is the functionality of the PC. An algorithm and LBP describe the binary pattern for neighbor pixels and generate pattern for the mapping. So proposed work will increase the true matching rate and decrease the false match rate with cryptography. This proposed method is appropriate for real-time application.