动态环境下基于改进LK光流的运动目标检测算法OA

Lucas-Kanade optical flow method,abbreviated as LK optical flow method,is one of the classic and commonly used moving object detection algorithms in the field of traditional image processing.The LK optical flow method has high accuracy,but the calculation of the optical flow field is relatively complex and susceptible to environmental noise,resulting in poor real-time performance.To address these issues,an image pyramid is constructed first,and then the LK optical flow method is used to calcu-late corner optical flow,and the optimal homography matrix is calculated by matching the feature points between two frames to com-pensate for the current image motion.The obtained optical flow field area is extracted through binarization and morphological opera-tions to obtain a rough rectangular area of the motion target,improving the target contour.Finally,different target regions are marked by using the eight neighborhood method to extract targets more accurately.The experimental results show that compared with the original algorithm,the improved algorithm has higher detection efficiency and good recognition effect,and can achieve continu-ous recognition and classification of moving objects.