Adaptive Dual Threshold Based Moving Target Detection Algorithm
Ke Liang, Yongmei Jiang, Meng Long, Guangming Liang · 2018 IEEE 4th Information Technology and Mechatronics Engineering Conference (ITOEC) · 2018
For the problems of inter-frame difference method and background subtraction method in the detection of moving objects, such as incomplete extraction target, easy to produce holes, and sensitivity to light and external interference, this paper proposes an improved target detection algorithm based on adaptive dual threshold. This algorithm adopts adaptive dual threshold to realize the binarization of frame difference image, and then uses post-processing such as morphology to eliminate false targets and connect fracture targets, and achieve more accurate extraction of the target. Experimental results show that the improved algorithm can overcome the effects of complex background, illumination changes, and motion disturbances, and can effectively guarantee the accuracy, efficiency and stability of moving target detection.