Hybrid Graphical Password Authentication Technique Based On GLCM And Canny Edge Detection

The process of verifying the identity of someone requesting access to a computer system, network, or piece of hardware is known as authentication. One of the most popular forms of authentication is the usage of alphanumeric passwords. Because human memory is limited, things can be forgotten. Moreover, password-recovery software can crack text-based passwords. An alternate kind of password authentication that uses visuals rather than text is called a graphic password. This paper aims to create an automatic click-point graphical password system based on hybrid techniques that offer quick processing times, large password space with high-security levels, and low storage requirements. The suggested method is predicated on GLCM texture features (homogeneity and entropy), and canny edge detection. Additionally, the creation of an Android application based on the AES method with a password recovery problem to carry out the suggested job at an allowed level. Several devices are used to test and evaluate the application. Every strategy performs admirably in terms of point generation processing time. Compared to the prior graphical password method, the new approach enhances password strength in terms of password space. Regarding the randomization test, all protocols demonstrated excellent performance and passed all applicable tests. That shows a large password space of about with large entropy and a fast login time of about (6 s).