采用序列图像的被动光学测速技术

Passive optical velocity measurement using image sequences

在获取无人机等飞行器的飞行速度时,自主、隐蔽性好的测量技术具有重要的应用价值。提出了一种利用光学相机拍摄的地面景象进行飞行器速度测量的技术。文中详细阐述了该项技术的工作原理,利用同一景物在两幅有重叠区域的图像中的位移量及间隔时间计算飞行器的运动速度。并且分析了其中关键的图像匹配技术,采用SIFT算法获得抗旋转、尺度变化的稳定特征点。重点对测速精度的各个影响因素进行了理论分析,给出了利用多帧图像进行测速提高测速精度的方法,综合分析了速度测量中的误差来源和各影响因素。理论研究和分析结果表明:利用光学图像进行速度测算的方案正确,精度能够满足飞行器速度测量的要求。

To acquire the flight speed of unmanned aerial vehicles, the measurement techniques of autonomous and good concealment are important. In this paper, a method of aircraft flight speed measurement using the images from optical camera was presented. The work principle of this method was particularly discussed. In two images that had overlapped area, the moving distance of the same scene and the interval time of the images were used to calculate the velocity. The image registration method was introduced, which was the key technology. Stable feature points were extracted by the SIFF algorithm, which were resistance to rotation and scale changes. The influence factors on measurement precision were analyzed particularly. The method of using multiple images to improve the precision was presented. The theoretical study and the analysis results show that the flight speed measurement method by images is correct, and the precision meets the velocity measuring needs.