论文部分内容阅读
通过研究建筑物结构中点、线基本结构模型,构建了相应的基本结构模型库。为了验证该模型库的有效性,利用模型库产生的仿真数据构造训练数据实现了一种在K-SVD框架下的建筑物屋顶LiDAR点云数据去噪方法。实验结果表明,所构建的建筑物模型库对建筑物点云数据具有非常好的表示能力,能够有效地应用于点云数据的去噪。
By studying the midpoint and line basic structure model of the building structure, the corresponding basic structure model library is constructed. In order to validate the validity of the model library, the training data is constructed by the simulation data generated by the model library to realize a LiDAR point cloud data denoising method for the roof of a building under the K-SVD framework. The experimental results show that the constructed model library has a very good ability to represent the point cloud data of buildings and can be effectively applied to the denoising of point cloud data.