A Method of Scene Matching Navigation in Urban Area Based on Shadow Matching

The scene matching based on the fixed reference image is difficult due to the difference in the photographing times and the variation of lighting conditions, and the projection positions of buildings are also different due to the variation of the camera position. In this paper, a novel scene matching navigation method using building shadows is proposed. Based on the lighting model, reference shadow maps are generated using accurate DSM data. A robust shadow detection algorithm is proposed for the aerial images. The results show that the method proposed in this paper overcomes the inconsistency problem caused by the variation of illumination, time and height of buildings, and achieves excellent matching performance.