因无人机机载激光雷达(Light detection and ranging,LiDAR)数据具有离散性,在生成数字高程模型(Digital elevation model,DEM)时需选择有效插值方法。以荒漠植被区为研究背景,使用零均值标准化方法归一化点云回波强度,利用肘方法确定...因无人机机载激光雷达(Light detection and ranging,LiDAR)数据具有离散性,在生成数字高程模型(Digital elevation model,DEM)时需选择有效插值方法。以荒漠植被区为研究背景,使用零均值标准化方法归一化点云回波强度,利用肘方法确定最佳聚类数目,采用K-means方法对点云强度值聚类得到地面点云。在此基础上,采用克里金(Kriging)方法插值抽稀率为20%和80%的地面点云数据,且将点云高程作为变量,建立RBF神经网络预测模型,并通过线性回归检验方法对模型进行精度分析,采用Delaunay三角网内插生成高精度DEM。结果表明:采用K-means方法实现最佳聚类数目为4的聚类,得到地面点云48722个,在点云较优抽稀率20%的情况下,径向基函数神经网络(Radical basis function neural network,RBFNN)训练时间为56s,点云高程预测的决定系数R2为0.887,均方根误差RMSE为0.168m。说明使用RBFNN对K-means聚类滤波得到的地面点云进行高程预测效果较好,可为基于点云构建高精度DEM提供参考。展开更多
Using the elbow function method to determine the number of classification groups of goods and applying the K-means algorithm to classify technical goods, this paper constructs a scientific classification framework of ...Using the elbow function method to determine the number of classification groups of goods and applying the K-means algorithm to classify technical goods, this paper constructs a scientific classification framework of commodity structure and then measures the changes in China ＂s import commodity structure. It is found, first, that from 2000 to 2012, the import commodity structure in China underwent structural change. Shares of non-agricultural primary products and high-technology products increased sigmficantly, whereas shares of low-medium-technology products and medium-high-technology products declined. Second, from 2000 to 2006, the overall technology level of China＇s imports declined, but then increased from 2007 to 2012. Therefore, the Chinese Government shouM increase the share of high-technology and extra- high-technology products in total imports to stimulate structural transformation in manufacturing industries to make full use of foreign advanced technology and to prevent stepping into a low-end import commodity structure.展开更多
文摘因无人机机载激光雷达(Light detection and ranging,LiDAR)数据具有离散性,在生成数字高程模型(Digital elevation model,DEM)时需选择有效插值方法。以荒漠植被区为研究背景,使用零均值标准化方法归一化点云回波强度,利用肘方法确定最佳聚类数目,采用K-means方法对点云强度值聚类得到地面点云。在此基础上,采用克里金(Kriging)方法插值抽稀率为20%和80%的地面点云数据,且将点云高程作为变量,建立RBF神经网络预测模型,并通过线性回归检验方法对模型进行精度分析,采用Delaunay三角网内插生成高精度DEM。结果表明:采用K-means方法实现最佳聚类数目为4的聚类,得到地面点云48722个,在点云较优抽稀率20%的情况下,径向基函数神经网络(Radical basis function neural network,RBFNN)训练时间为56s,点云高程预测的决定系数R2为0.887,均方根误差RMSE为0.168m。说明使用RBFNN对K-means聚类滤波得到的地面点云进行高程预测效果较好,可为基于点云构建高精度DEM提供参考。
基金This paper is supported by the National Social Science Foundation （No. 14ZDA082 ）, the National Natural Science Foundation of China （No. 71473020）, the Social Science Foundation of Education Ministry of China （No. 14YJA790058） and the Beijing Social Science Foundation （No. 15JGA005）.
文摘Using the elbow function method to determine the number of classification groups of goods and applying the K-means algorithm to classify technical goods, this paper constructs a scientific classification framework of commodity structure and then measures the changes in China ＂s import commodity structure. It is found, first, that from 2000 to 2012, the import commodity structure in China underwent structural change. Shares of non-agricultural primary products and high-technology products increased sigmficantly, whereas shares of low-medium-technology products and medium-high-technology products declined. Second, from 2000 to 2006, the overall technology level of China＇s imports declined, but then increased from 2007 to 2012. Therefore, the Chinese Government shouM increase the share of high-technology and extra- high-technology products in total imports to stimulate structural transformation in manufacturing industries to make full use of foreign advanced technology and to prevent stepping into a low-end import commodity structure.
