舰船电子工程2026，Vol.46Issue(3)：45-50,6.DOI:10.3969/j.issn.1672-9730.2026.03.010
动态环境下基于改进LK光流的运动目标检测算法
Moving Target Detection Algorithm Based on Improved LK Optical Flow in Dynamic Environment
摘要
Abstract
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.关键词
运动目标检测/LK光流法/背景补偿/动态背景/单应矩阵Key words
moving target detection/LK optical flow method/background compensation/dynamic background/single-re-sponse matrix分类
信息技术与安全科学引用本文复制引用
钱奕舟,王在俊,高耀文,王雪..动态环境下基于改进LK光流的运动目标检测算法[J].舰船电子工程,2026,46(3):45-50,6.基金项目
民航飞行技术与飞行安全重点实验室开放基金项目"基于深度强化学习的多无人机博弈对抗技术研究"(编号:FZ2020KF07) （编号:FZ2020KF07）
科研基地创新创业基地建设项目(编号:E2023065) （编号:E2023065）
2023年中国民用航空飞行学院大学生创新创业训练计划项目(编号:S202310624230) （编号:S202310624230）
中国民用航空飞行学院研究生创新项目(编号:24CAFUC10175)资助. （编号:24CAFUC10175）
