Considering the shadow areas of the planet surface can be seen as the landmark in the process of guidance, navigation and control(GNC) and hazard avoidance when the spacecraft soft lands on the small body, a method of the shadow areas tracking is proposed. Firstly, with the attitude estimates that are supplied by the attitude-measuring instruments, the two consecutive frames of image are rectified to the same attitude, and the shadow areas are detected by using the proposed binary algorithm. Then, the high correlation points of the shadow areas are found by using the template matching algorithm, and the feature vectors of the high correlation points are extracted to find the most similar point. Finally, the false matches are removed by using of random sample consensus(RANSAC). Two groups of sequence images of 433 Eros are utilized to teste the performance of the proposed method, and the results show that the method can be used in the real sequence images of planet.