一种适用于空间非合作目标的角点检测算法

随着在轨失效航天器数量不断增加,针对这类空间非合作目标的在轨服务需求日益迫切。由于在轨的非合作目标飞行器上无法布置合作标志器,通过提取非合作目标上的关键轮廓角点对其进行跟踪。本文针对空间非合作目标所具有的特性,传统Harris算子对运动非合作目标角点的检测中存在大量的伪角点,提出一种基于Canny边缘检测的Shi-Tomasi角点检测方法,即对影像先进行滤波、二值化及形态学处理,其次利用Canny边缘检测获得其轮廓后再进行Shi-Tomasi角点检测。实验结果表明,该算法能有效地解决深空中光照条件等因素造成的角点难以检测问题以及减少伪角点,对空间非合作目标的关键角点提取有较好地检测精度和准确性。

A key feature point detection algorithm for non-cooperative target in space

With the increase of the number of spacecraft in orbit,the demand for the service of this kind of non-cooperative target in space is becoming more and more urgent.Due to the problem that orbit non-cooperative target aircraft cannot be arranged cooperative markers,the key outline corners extracted from non-cooperative target can track the target.Aiming at the characteristics of space non-cooperative targets, the traditional Harris algorithm has a large number of pseudo-corner points in motion detection of non-cooperative targets.To solve the problem,the paper proposes a method combined Canny edge detection with Shi-Tomasi corner detection.First,the process for images includes filtering,binarization and morphological processing,then the Canny edge detection is used to obtain the contour and combine with Shi-Tomasi corner detection to obtain the result.The experimental result shows that this algorithm can effectively solve the difficult detection problems and pseudo-corner points in deep space due to the lighting conditions,of w hich the detection has better precision and accuracy.