Covid Protocol Enabled Dual Mechanism for Authentication

Abstract: In the Modern day, with such advancement in technology, the world is going touchless for security, driving, and many such activities. In this paper, we tried creating an advanced security system for face and fingerprint recognition by touchless means. Though much efforts can be put to further increase the efficacy and accuracy of the prototype, but this is just a beginning. First, face image will be captured and matched with the already present images in the database using a pre-trained model based on Deep Neural Network (DNN) which takes 128 different measurements for face identification. Age and Gender prediction model and prototype is also being used to predict gender and probable age range. Second, if step 1 is successful, the prototype asks for fingerprint image using second camera. The image is then processed to extract fingerprint lines similar to minutiae extraction with help of Gaussian Adaptive Threshold Algorithm. The lines are then sharpened, blurred and enhanced. The final processed image is then matched with the already saved images in database. The points in the images are matched using SIFT algorithm which enables to detect even if the image is at different angle, size, blurred or any other discrepancy. Lastly if both Face and Fingerprint are matched with same individual data, the prototype grants the access.