一个小而有效的程序来执行的K -近邻搜索算法，此算法利用JIT 理论加速循环，比向量化有效解决了大量数据的精度问题。甚至比kd-tree效果要佳-A small and effective procedures to implement the K- nearest neighbor search algorithm, this algorithm uses JIT acceleration cycle theory, than to solve a lot of data to quantif

kd-tree 是一种常用的数据结构，它在搜索最近点中有普遍用处。该文件是一个用纯matlab写成的源代码。-kd-tree is a commonly used data structure, it is commonly used in searches for the nearest point. The file is a pure source code written in matlab.

基于kd-tree搜索算法，利用局部邻域的平面拟合，检测三维点云数据的法线方向- Estimates the normals of a sparse set of n 3d points by using a set of the closest neighbours to approximate a plane.