Point cloud hole repair based on weighted combination algorithm
In order to repair point cloud holes in UAV aerial survey data,LS-SVM algorithm and BP optimized by GA(BP-GA)are linearly combined to construct a weighted combination model for hole repair in scattered point cloud data.By combining the errors of the two methods,a weighted combination model related to the er-rors of two repair methods is established,and it will repair the point cloud holes.The residual and internal and external accuracy of the repair results of weighted combination model,LS-SVM and BP-GA are analyzed.Compared with the results of the two single repair methods,the point cloud repair results obtained by the weigh-ted combination model have higher internal and external coincidence accuracy and higher stability,which pro-vides an effective hole repair method for the point cloud data obtained by UAV.
point cloud holeLS-SVMBP neural networkweighted combinationhole repair
国家科技期刊平台
NSTL
万方数据
