k近邻空间插值算法优化研究

An Algorithmic Optimization Study of k-nearest Neighbor Space Interpolation

在k近邻空间插值中,如果能减少近邻点的搜索次数,可进一步提高空间插值的性能。引入k近邻距离阈值的概念和计算方法,并以该阈值为基础,发展了k+M优化算法。其算法核心是在空间插值过程中,获取初始栅格的k+M近邻点集,计算k+M近邻距离阈值。若从初始栅格向右移动至其他栅格的距离小于该阈值,则直接利用初始栅格的近邻点集进行空间插值。实验证明,该算法相对于每个栅格均搜索近邻点的算法,性能有明显的提升。

For the k-nearest-neighbor spatial interpolation algorithm,the performance of spatial interpolation can be further improved if the number of searches for nearest-neighbor points can be reduced.In this paper,the concept and calculation method of the k-nearest neighbor distance threshold is proposed first,and based on it the k+M optimization algorithm is developed.The key part of the k+M optimization algorithm is obtaining the k+M nearest neighbor points of the initial grid in the process of spatial interpolation,and calculating the k+M nearest neighbor distance threshold.If the moving distance of the initial grid is less than the set threshold,the spatial interpolation can be performed directly using the nearest neighbor points of the initial grid.Experimentally,compared with the algorithm of searching the nearest neighbor points for each raster,the performance of the algorithm used in this paper is significantly improved and its research results have certain theoretical significance and practical value.