Shadow–highlight feature matching automatic small crater recognition using high-resolution digital orthophoto map from Chang’E Missions

This paper introduces a new method of small lunar craters’ automatic identification, using digital orthophoto map (DOM) data. The core of the approach is the fact that the lunar exploration DOM data reveal contrasting highlight and shadow characteristics of small craters under sunlight irradiation. This research effort combines image processing and mathematical modeling. Overall it proposes a new planetary data processing approach, to segment and extract the highlight and shadow regions of small craters, using the image gray frequency (IGF) statistical method. IGF can also be applied to identify the coupling relationships between small craters’ shape and their relative features. This paper presents the highlight and shadow pair matching (HSPM) model which manages to perform high-precision automatic recognition of small lunar craters. Testing was performed using the DOM data of Chang’E-2 (CE-2). The results have shown that the proposed method has a high level of successful detection rate. The proposed methodology that uses DOM data can complement the drawbacks of the digital elevation model (DEM) that has a relatively high false detection rate. A hybrid fusion model (FUM) that combines both DOM and DEM data, was carried out to simultaneously identify small, medium, and large-sized craters. It has been proven that the FUM generally shows stronger recognition ability compared to previous approaches and it can be adapted for high precision identification of craters on the whole lunar surface. The results meet the requirements for a reliable and accurate exploration of the Moon and the planets.