Abstract: In order to solve the problem of low vegetation penetration rate of UAV photogrammetry and the difficulty in extracting ground objects from UAV LiDAR(Light Detection and Ranging) point cloud, a high-precision river topographic mapping method integrating UAV photogrammetry and LiDAR was proposed.In this method, DEM(Digital Elevation Model) with high resolution was extracted from the LiDAR point cloud by ground filtering, the true color 3D point cloud was generated through the fusion of LiDAR point cloud and DOM(Digital Orthophoto Map),the real scene 3D model was reconstructed by the fusion of LiDAR point cloud and digital images, and then the accurate mapping of ground features was realized.The typical application results showed that this method could give full play to the complementary advantages of the two technologies, and could solve the problems of high-density vegetation covered geomorphic surveying and ground features extraction of diverse buildings in the river bank area of the lower Yangtze River.The data acquisition and processing were efficient, the products were intuitive, fine and diversified, and the accuracy could meet the requirements for 1∶1 000 large-scale topographic mapping.