A Lunar Terrain Reconstruction Method Using Long Base-line Stereo Vision

In order to make lunar lander land on the surface of moon autonomously and safely, the terrain of landing region should be rebuilt, and the description of landing site terrain should be generated. A long base-line stereo vision 3D reconstruction method through motion is proposed. Aimed at the problems of long base-line stereo vision, a series of algorithms are designed to select and match feature points, to estimate the relative rotation and translation between different positions, to rectify stereo image pair, to get the dense disparity image, to perform 3D reconstruction methods through disparity images, and to generate the DEM (digital elevation map) of the landing region. According to the characteristic of lunar terrain, a software simulation test-bed is built, on which the algorithms are tested. The simulations show that this method can be effectively used in 3D reconstruction of lunar terrain.