Hierarchical classification for the topography analysis of Asteroid (4179) Toutatis from the Chang'E-2 images

High spatial resolution images of the near-Earth Asteroid (4179) Toutatis are provided by a successful flyby of the Chang'E-2 spacecraft. These optical images give us a chance to closely observe the surface of this asteroid. However, some local low-contrast regions in the Chang'E-2 images limit the accuracy of the topography recognition. To solve this problem, a hierarchical classification method is suggested and developed to assist the topography analysis for Toutatis based on the Chang'E-2 optical images. The proposed method first classifies the image at both the macro level and the micro level, respectively. Then, topography of Toutatis can be explored by using the hierarchical classification result. Experimental results demonstrate that the method cannot only provide a new perspective to analyze topography objects reported previously, but also reveal some new characteristics at low-contrast regions. Namely, two new topographic characteristics are revealed, one is a connection region with a particular spectral value locating at the corner of the large lobe, and the other is an object seeming like a fixed star at the background of the images.