Abstract: Safety is a major concern in our lives today. Whether within an organization or in a restricted area; With the growing need for security measures in everyday life, biometrics has become a hot topic of research targeting its potential value in personal identification. This is because biometric systems are classified as more secure than other security systems. This paper focuses on iris and fingerprint as one of the best biometric features for identity management. Iris recognition possesses properties that make it a quintessential biometric system. The point of this venture is to distinguish an individual without a blunder, burning through less time, and keep away from mistakes in confined regions. The recognition of Iris for dealing with Indian weapons is the most powerful innovation related to the security of our country. The method used for fingerprint authentication divides the identification into stages and eliminates many fake fingerprints at different stages. This saves a lot of time by maintaining a high recognition rate. Although Minutiae-based technology is widely used, it is difficult to extract features if the fingerprint image quality is poor. This paper aims to understand how the characteristics of fingerprints can be inferred. Significant advances in this field show that iris and fingerprint biometrics still require fast, real-time, reliable, and powerful algorithms for higher recognition.

Keywords: CNN(Convolutional Neural Networks), deep neural network, drowsiness, python, Iris, Hough Transform, Daugman Method, Fingerprint, Minutiae, Image Processing.

| DOI: 10.17148/IARJSET.2023.10216