Graphical password by image segmentation

Abstract

Image segmentation is a very important technique to analyze an image and partitioning it throw detecting its pixels and separating them depending on their color, intensity, or texture. Object segmentation methods have helped in many computer vision tasks, such as scene representation& interpretation, object tracking in videos, medical field, and many more. It is a crucial component in many image analysis and computer vision areas. The main purpose of this project is to make a graphical password based on the image segmentation. The idea of this topic is to enter to the system many images which will be divided into different segments. When the user wants to open the system, the segments appear on the screen but randomly located. To pass the code the user must click on some chosen segments. There are many methods of image segmentation. In this project, we used segmentation by the normalized cuts method in combination with the spectral clustering and the k-means clustering method.