一种结合灰度特征的月面撞击坑提取方法

摘要

In order to accomplish the task of positioning and navigation of the detector using the impact crater,an automatic extraction method of the crater based on image gray feature is proposed.Firstly,the illumination direction of the image is calculated by using the method of gray gradient statistics.Then,the image is divided into three regions,which are the shadow region,the highlight region and the other regions by using the double threshold segmentation method.In the segmentation result,the shadow area is the seed point,and the corresponding highlight area is searched constrain by the light direction condition,and then the response area search is completed.Finally,the shape of craters is obtained by edge fitting.Experiments were carried out using the lunar image and LRO image.The results show that for different illumination images,the proposed method has higher searching rate and anti noise,and this has laid a good foundation for the further research of the landing mission and the high precision reconstruction of the landing zone in the deep space exploration.%针对利用撞击坑辅助完成探测器的定位、导航任务的问题,提出了一种基于图像灰度特征的撞击坑自动提取方法.首先采用灰度梯度统计的方法获取影像的光照方向,然后采用双阈值分割方法将影像分为阴影区、高亮区和其他3种区域,并以分割结果中的阴影区为种子点,光照方向作为约束条件,搜索其对应的高亮区,完成响应区域搜索,最终通过边缘拟合得到撞击坑形态.采用我国探月影像和美国LRO影像进行了实验,结果表明对于不同光照条件下的影像,该文的方法都具有较高的搜索率和抗噪性,为后续深空探测任务的着陆任务及着陆区的高精度重建等研究打下良好的基础.