Rock detection based on 2D maximum entropy thresholding segmentation and ellipse fitting

In order to make lunar lander autonomous safely on the moon, the potential hazard should be detected. Effective detection of rocks will also provide information for choosing a better landing region which is suitable for exploration. The 2D maximum entropy thresholding method is proposed to segment the shade area from the image, and then the ellipse fitting method is used to estimate the sizes and locations of the rocks. The simulation results demonstrate that these methods can effectively detect the rocks in an image.