可见光和红外热像仪融合下运动目标跟踪方法

打开文本图片集

关键词： 运动目标跟踪； 可见光； 红外热像仪； SURF； HOG； 图像配准中图分类号：TN219⁃34；TP391 文献标识码： A 文章编号：1004⁃373X（2025）12⁃0111⁃04

Abstract：In complex image scenes， such as the presence of noise interference， partial occlusion of targets， or complex background textures， the calculated moments may be affected， resulting in inaccurate feature point centers and affecting the tracking effect of moving targets. A method of motion target tracking fusing visible light and infrared thermal imager is proposed. The infrared thermal imager is used to obtain the initial moving target. During the motion of the infrared thermal imager， as the captured image may be distorted due to camera movement， rotation， or target object motion， the optical motion model is introduced. For the corrected image， the feature point centers are determined by calculating the moments of the image blocks， and the nearest neighbor ratio matching strategy is adopted to determine the effectiveness of the matching pairs by comparing the distance ratios between the feature points and their nearest and second nearest neighbors. By improving the combination of SURF and HOG， the infrared image and visible light image of the target are registered and detected， completing the registration and fusion of the infrared and visible light image pairs. After registration and fusion of infrared and visible light images， the motion target tracking is realized by means of multi⁃feature fusion correlation filtering methods. The experimental results show that the proposed method has higher accuracy in tracking moving targets and better practical application effects.

Keywords：motion target tracking; visible light; infrared thermal imager; SURF; HOG; image registration

在现代视觉监控与跟踪领域，运动目标的精确识别与持续跟踪是一项至关重要的任务，尤其在复杂多变的环境条件下，如夜晚、大雾、烟雾等低能见度场景[1⁃2]。（剩余5348字）