摘要
针对在复杂环境中对小目标难以跟踪的问题,提出了一种改进金字塔Lucas-Kanade(LK)光流法用于检测追踪小目标.通过将Shi-Tomasi角点检测和金字塔LK光流法结合,对复杂环境中的光流场进行筛选,然后对运动目标进行角点检测,最后通过中值滤波去除视频中的噪声干扰,使算法精确度和效果更好.本研究通过大量实验,在加入中值滤波后将金字塔LK光流法的金字塔层数和迭代次数进行更改,使光流法更加适用于本研究.同时将改进后金字塔LK光流法与现有金字塔LK光流法进行对比,可以明显看出改进后的算法具有更好的鲁棒性和准确率.改进光流法在复杂环境中对小目标进行检测具有比其他方法更好的跟踪检测效果.
Aiming to resolve the problem of tracking small targets in complex environments,an improved pyramid Lucas-Kanade(LK)optical flow method has been proposed to detect and track small targets.By combining Shi-Tomasi corner detection with the pyramid LK optical flow method,the optical flow field in complex environments is filtered,and the corner detection is performed on moving targets.Finally,a median filter is used to remove noise interference from the video,making the algorithm more accurate and effective.In this study,the number of pyramid layers and iterations of the pyramid LK optical flow method was varied through numerous experiments after the addition of a median filter,making the optical flow method more suitable for this investigation.At the same time,comparing the improved pyramid LK optical flow method with the existing pyramid LK optical flow method,it can be clearly seen that the improved algorithm has better robustness and accuracy.The improved optical flow method achieved superior tracking and detection performance for small targets in complex environments compared to other approaches.
作者
邹林
ZOU Lin(Dali State Energy Conservation Monitoring Center,Dali 671099,China)
出处
《西安文理学院学报(自然科学版)》
2025年第3期48-55,共8页
Journal of Xi’an University(Natural Science Edition)
