基于二维最大熵阈值分割的坑识别方法

Hole Detection Based on 2D Maximum Entropy Threshold Segmentation

月球环境感知是未来月球探测车在月球表面实现自主导航一个非常重要的环节,坑是月球表面存在较多的一种地理特征,将成为月球探测车导航的障碍物。文章从模式识别的分类理论出发,提出了一种基于二维最大熵阈值分割的地球环境坑检测方法,为将来月球探测车的环境感知模块提供技术参考。首先利用二维直方图熵最大化阈值分割方法将图像二值化,然后结合形态学滤波算法和阈值面积消除法滤除一些噪声,最后对所得到的二值图像进行轮廓提取并在原图上显示,实现了对部分场景下坑的检测。

The environment appereeption is a very importation for the moon vehiele to navigation in the moon.The hole,which is the main geography character on moon,will be a obstacle to the navigation of moon vehicle.This paper introduces an earth environment hole detection method based on 2D maximum entropy threshold segmentation and gets good effect.First the 2D maximum entropy threshold method is used to segment the hole area from the image,then the morphological filter and area threshold removing method is applied to filter the small area and the noise,Finally the hole contour is abstracted from the binary image and displayed on the original image.