Particle Filter and CAMShift Approach for Motion Detection: A Comparative Study

Object Detection and Motion Detection are the wide research areas of computer vision. Various methods are available for performing these tasks. Here, two algorithms Particle Filter and CAMShift are used. The parameters such as intensity, elapsed time, track loss rate etc. are examined for both algorithms. Particle filter is a technique for implementing recursive Bayesian filter by Monte Carlo sampling. The posterior density is represented by a set of random particles with associated weights and the estimates are computed based on these samples and weights. CAMShift algorithm is improvement over the MeanShift algorithm. It uses the color histogram model of the target to convert the image into the color probability distribution map, initialize the size and position of a search window, and adaptively adjust the position and the size of the search window according to the results obtained from the last frame, thus locating the center position of the target in the current image.