Journal of Astronautics ›› 2016, Vol. 37 ›› Issue (2): 169-174.doi: 10.3873/j.issn.1000-1328.2016.02.005

Previous Articles     Next Articles

Egress Mechanism Recognition and Slope Measurement for Mars Exploration

LI Ying, YE Pei jian, PENG Jing, DU Ying   

  1. China Academy of Space Technology, Beijing 100094,China
  • Received:2015-03-25 Revised:2015-05-12 Online:2016-02-15 Published:2016-02-25


A method for recognition and location of egress mechanism and its slope measurement based on a stereo vision technique is proposed for Mars exploration in this paper. First, feature matching is used for egress mechanism recognition and location. Taking account of image affine transform caused by the change of view angle of rover navigation camera and the slope of egress mechanism, the effect of lighting condition on the Mars surface and computer ability of the Mars rover, combined with the own attribute of egress mechanism, a method for recognition and location is proposed based on feature fusion of region feature and Blob feature.The speeded up robust features (SURF) algorithm is used to detect local invariant feature, and the maximally stable extremal regious (MSER) algorithm is used to detect the maximally stable extremal regions.Both features are combined and described by using SURF descriptor. Meanwhile, the M-estmator sample consensus (MSAC) algorithm is applied to compute the transformation matrix of matching points and realize the preliminary location of the egress mechanism in the scene image. Then, the slope of egress mechanism is computed by locating the artificial marks accurately and 3D reconstruction. The experimental results indicate that the method proposed is robust to the change of both view angle and lighting condition. The number of valid feature points extracted and the computer time are better than the performance of scale invariant feature transform (SIFT), and meets the requirements of Mars exploration system.

Key words: Mars exploration, Stereo vision, Target recognition, Rover egress, Robust feature

CLC Number: