Rectangular Object Tracking Based on Standard Hough Transform

Object tracking in computer vision is important. Numerous research papers have been published about this problem. Few references relating to tracking shaped objects via the Hough Transforms exist. This paper provides a method based on the Standard Hough Transform to track rectangular objects. Our method works on edge images obtained by applying the Canny edge detector to the source image. The rectangular object in the edge image is then extracted based on the Standard Hough Transform and related conditions of the rectangle. Afterwards, we shift the region of interest to the new position of the tracked object and continue to detect the rectangle in successive frames. Our technique can be run in real-time due to its low computation time and simplicity.