Abstarct: Object tracking can be defined as displaying an object's location changes and tracking it in a series of video images. Most of the practical tracking applications require a real-time algorithm. Also, it's notable that being real-time is more significant in security and military applications and has a remarkable effect on the performance of these systems. The importance of the object tracking process is due to the fact that while in the recent years we’ve witnessed the use of cameras with high resolution in tracking systems; in most tracking methods, this issue causes the amount of information in each frxame that should be processed to increase. In addition, considering the growing progress of technology and the increasing demands from smart systems, processing flows have become more complex in recent years. Therefore, it’s impossible to achieve real-time processing with only software operations, and the use of hardware capabilities is necessary in order to achieve the desired results. In this thesis, in order to perform real-time object tracking, we use parallel processing by the graphical processing unit besides the CPU, assuming a stationary camera is used. We implement some of the existing methods in real-time object tracking like frxame difference, three-frxame difference and particle filter on GPU and CPU. Finally, with comparison of the performance of the parallel algorithms on videos with various dimensions and complexities, we show that parallelization of the object tracking process using GPU besides the CPU in order to achieve real-time tracking would be an appropriate approach. Obtained results show that processing time improves more than thousand times if the problem of too much data transmission delay is solved. For example, in the case of particle filter with 1200 particles, running the algorithm on GPU improves the speed by a factor of 1130 compared to the CPU. Relative to video dimensions using GPU for frxame difference and three-frxame difference results in speed increase by 1390 and 548 times respectively.