Abstract: To explore the fusion application of multi-source heterogeneous data in shallow water river section mapping technology with abundant aquatic plants and green trees,and to address the difficulty of extracting river sections with high precision and efficiency,a method is proposed to use high-precision DEM of river channels to extract cross-sectional scatter points for cross-sectional processing.This method uses optical satellite remote sensing images to pre-layout the cross-sectional line of the river channel.A UAV is used for low altitude aerial photography and point cloud scanning measurement of the river channel.For smooth and flat water areas without grass,a single beam depth sounder is used on an unmanned ship for water depth measurement.For shallow water areas with more grass in the river,GNSS RTK is used for underwater terrain point acquisition.Using processed data fusion to generate a water land integrated DEM with a resolution of 0.2 m,transverse scatter points with a spacing of 200 m are extracted.Any point is selected from the transverse scatter points with three-dimensional coordinates as the endpoint to calculate the station number and offset.Based on the endpoint coordinates and the perpendicular line drawn from the endpoint to the centerline of the river channel,the observation direction of the cross-section is considered.The offset distance between the endpoint and the transverse scatter points is calculated to draw the cross-section map.The total station is used to conduct high-precision inspection of the surveying results,and the error of 168 cross-sectional points is calculated to be 0.065 m.The quality level of the measurement results is evaluated,and the quality element score is 94.9 points,indicating that the quality level is rated as excellent.

Key words: UAV, unmanned ship, GNSS RTK, DEM, sectional measurement