Human detection and tracking from image sequences using background subtraction technique

67 p.In this dissertation, we present a histogram based statistical background subtraction method for human detection and tracking from a static background scene. Experimental tests are performed using the prerecorded videos taken in outdoor environments. In our proposed algorithm, histogram based image segmentation is used for background subtraction. By using the histogram results, we have identified each pixel or super pixel in each current frame whether it is a background pixel or a foreground pixel. The results of the experiments suggest that what kinds of background are suitable for our proposed algorithm.Master of Science (Computer Control and Automation