【摘要】 常用的运动目标检测算法有背景减除法、帧间差分法和光流法,针对背景减除法的背景模型需要实时更新,帧间差分法检测到的目标不完整,本文提出将Lucas-Kanade光流法与最大类间方差的图像分割法相结合的算法,即首先对连续两帧图像进行Lucas-Kanade光流计算,再对其进行最大类间方差图像分割,将光流不连续的区域视为运动目标,光流连续的区域视为背景,最后进行形态学处理,完成运动目标的检测过程。通过Matlab实验仿真,验证了本文所提算法能提取更完整的运动目标,检测效果较好。更多还原

【Abstract】 The common methods of moving target detection are background subtraction,frame difference and optical flow method.Because the background model of background subtraction needs to be updated in real time and the detected target is incomplete by the frame difference method,this paper proposes an algorithm combining the Lucas-Kanade optical flow method with the image segmentation method of maximum inter class variance. Firstly,calculate two consecutive frames images by the Lucas-Kanade optical flow algorithm; secondly,process the images by segmentation method of maximum inter class variance. So the discontinuous region of the optical flow is regarded as the moving target and the continuous region of the optical flow is regarded as the background. Finally,the morphological processing is performed to process images. This is the whole detection process of the moving target. The simulation results of Matlab show that the proposed algorithm can extract more complete moving target and the detection effect is better.更多还原