Method for rapid construction of urban three-dimensional real scene model
The traditional point cloud filtering algorithm uses a one-sided scene and requires manual intervention.In addition,its adaptive performance of the threshold is weak,which leads to a poor construction effect of the three-dimensional(3D)model to a certain extent.Therefore,the study proposed an adaptive threshold improvement filtering algorithm based on mixed least squares surface fitting.Algorithm validation shows that the algorithm proposed in the study can achieve high-precision filter processing in urban,rural,and forest scenes,and the average error of the filtering effect of the sample data is only 7.04%,a reduction of 12.22%-72.69%compared with other algorithms.The filtering effect of this algorithm is more stable in different scenes and is superior in discontinuous geographic areas and urban scenes with many buildings.A comparison of modeling effects shows that the accuracy and geometric structure processing of the 3D model filtered by the proposed algorithm are superior,and the generated 3D model is closer to the real building.The results show that the proposed adaptive threshold improvement filtering algorithm has better filtering accuracy for different scenes and can effectively retain the geographical features and terrain details of the real scene.The constructed 3D model has better accuracy and detail processing effect,which is conducive to improving the construction efficiency of urban 3D real scene models.
万方数据
维普
