基于红外图像处理的移动目标识别跟踪算法

Moving target recognition and tracking algorithm based on infrared image processing

为解决红外热像仪在战场环境多变、背景复杂、目标运动速度过快等情况下难以对目标进行识别与跟踪的问题,本文提出了一种基于红外图像处理的移动目标识别跟踪算法。该算法主要分为移动目标的检测和跟踪两部分。首先采用高斯背景建模方法来提取运动目标,其次应用Kalman滤波来进行运动目标的定位与跟踪,最后通过匈牙利算法来判定目标点的归属。实验结果表明,在MOT基准挑战中的评价指标上,本文算法相比于Sort跟踪器将MOTA提升了1.4%,MT提升了11.9%,ML降低了6.5%,IDs降低了33.3%,FN降低了13.6%,提高了跟踪准确性,成功实现了对多个移动目标的跟踪。本文算法在军事领域具有广泛的应用前景。

To address the problem of difficulty in recognizing and tracking targets using infrared thermal imaging cameras in a complex and rapidly changing battlefield environment,a moving target recognition and tracking algorithm based on infrared image processing is proposedin this paper.The algorithm mainly consists of two parts:detection and tracking of mobile targets.Firstly,the Gaussian background modeling method is used to extract moving targets.Secondly,Kalman filtering is applied to locate and track moving targets,and finally,the Hungarian algorithm is used to determine the attribution of target points.The experimental results show that compared with the SORT tracker,the proposed algorithm improves MOTA by 1.4%,MT by 11.9%,reduces ML by 6.5%,IDs by 33.3%,and FN by 13.6%,which improves the tracking accuracy and successfully realizes the tracking of multiple moving targets.The algorithm in this paper has broad application prospects in military fields.