带有地理编码的光学视频卫星物方稳像方法

A Geo‑Coded Stabilization Approach for Optical Video Satellites in Object Space

  • 摘要: 发现运动目标并获得运动目标的地理位置与运动速度,是光学视频卫星非常重要的一个高精度应用。现有光学卫星视频稳像方法虽然可以获得稳定流畅的视频数据,但视频数据中缺少几何信息,难以满足这一应用需求。针对该问题,提出一种带有地理编码的光学视频卫星物方稳像方法。该方法充分利用卫星视频帧的定向参数信息,构建基于视频帧定向模型的帧间运动模型,并通过辅帧定向参数的运动补偿,实现视频序列帧间几何一致性,再在物方空间对视频序列进行地理编码,生成带有地理编码的流畅视频数据。经对珠海一号卫星视频数据的试验结果表明,该方法可以有效消除卫星平台抖动和卫星定轨测姿等误差对卫星视频帧间几何精度的影响,获得的帧间几何精度与视频稳像精度均优于0.3像素。

     

    Abstract:
      Objectives  Discovering a moving target and obtaining its geographic position and moving speed is a very important and highly accurate application for optical video satellites. Generally, existing stabilization approaches for optical video satellites can obtain stable and smooth videos. However, these stabilized satellite videos often have no geometric information, which makes them difficult to meet the application requirements. To solve this problem, a geo-coded stabilization approach for optical video satellites in object space is proposed.
      Methods  The proposed approach takes full advantage of the orientation parameter information of satellite video frames. Firstly, an interframe motion model based on the frame orientation model is established. Secondly, orientation parameters of the auxiliary frame are motion-compensated to achieve a geometric consistency between the adjacent video frames. Finally, all the video frames are geo-coded in object space, and a geo-coded and smooth video can be generated.
      Results and Conclusions  The experimental results of a set of Zhuhai-1 satellite video frames show that the proposed approach can effectively eliminate the influence of satellite jitter errors and satellite position and attitude errors. Both the interframe geometric accuracy and the video stabilization accuracy achieved by the proposed approach reach better than 0.3 pixels.

     

/

返回文章
返回