Binary Vision Algorithms in Java

The capabilities of Java as a highly portable, multi-threaded general purpose programming language can be effectively utilized to provide an efficient framework for learning, coding, visualizing and demonstrating the fundamental concepts behind binary vision algorithms. Methods whose implementations are described in this paper include connected component labeling, boundary following algorithm, image feature extraction, and image thinning algorithm. The concept of a pseudo-screen in introduced to generate the display of a scaled pixel grid for a highly magnified view of the pixel level operations and the intermediate stages in recursive computing. The graphics and the user interface classes of Java together with the thread class are used to create methods necessary for interactive input and update of images and also for rendering the output on any Java enabled browser