Research on accuracy optimization of 3D laser point cloud modeling with GNSS elevation constraints

This study proposes a 3D laser point cloud modeling method integrating GNSS elevation constraints. Building upon traditional TIN-DEM reconstruction, it introduces a collaborative algorithm combining hard and soft constraints to construct a weighted least squares fusion model, achieving precise alignment between point clouds and GNSS elevation data. Using a 3.2 km² dataset from the coastal hilly region of Shandong Province as an example, the integrated approach reduces elevation RMSE from 13.8 cm to 6.4 cm, lowers outlier rates to 1.2%, and decreases surface roughness by an average of 35%. It demonstrates enhanced stability and applicability in areas with complex terrain features and steep slopes, providing reliable technical support for high-precision topographic surveying.