Moving Target Detection Algorithm Combining Three-Frame Difference and Hough

Difference and Hough Mingming Liu1, Dong Pei 1,2, *, Haoxiang Sun 1, Ju Liu 1, Donghui Zhu 1 1College of Physics and Electronic Engineering, Northwest Normal University, Lanzhou Gansu 730030, China; 2Engineering Research Center of Gansu of Gansu Province for Intelligence Information Technology and Application, Lanzhou Gansu 730030, China. *peidong@nwnu.edu.cn Abstract. Aiming at the three-frame difference method, the moving target detection will be misdetected due to the influence of noise. A moving target detection algorithm with improved threeframe difference combined with Hough is proposed. Firstly, the background difference method based on local adaptive threshold is used to reduce the influence of light and noise on moving target detection. Secondly, the differential result of three-frame difference method is introduced to improve the sensitivity of detection. Finally, the two differential results are combined and combined. The Hough transform detects the moving target. The experimental results show that the algorithm can effectively suppress noise and quickly and accurately identify moving targets, which meets the needs of real-time detection.