kdtree module for C++

1. Copy kdtree.hpp and node.hpp to your project.
2. Write #include "kdtree.hpp" in your code.

This module uses namespace kdtree.
Write using namespace kdtree; if necessary.

• kdtree : A class having a pointer to the root node of the tree.
• node : A class representing a node of the tree.

Generally you don't have to create a node directly, but it might be good to know how to create a node and access to its data.

node is a template class and you can use any classes having the member x and y.

node<cv::Point> *node = new node<cv::Point>(cv::Point(10, 0));

The member point represents the point of the node.

cout << node->point << endl; // [10, 0]

Read node.hpp if you want to know about other methods and member variables.

kdtree is a template class and you can use any classes having the member x and y.
Following example uses cv::Point of OpenCV as a point of the tree.

vector<cv::Point> points = {
    cv::Point(10, 0),
    cv::Point(20, 0),
    cv::Point(40, 0),
    cv::Point(80, 0),
    cv::Point(160, 0)
};

kdtree<cv::Point> *tree = new kdtree<cv::Point>(points);

Read kdtree.hpp if you want to know about other methods and member variables.

Just delete the instance of the tree.

delete tree;

Use nearest().

In this example the result is the node having cv::Point(40, 0).

node<cv::Point> *nearest_neighbor = tree->nearest(cv::Point(50, 0));

In this example the result is the node having cv::Point(80, 0).

node<cv::Point> *nearest_neighbor = tree->nearest(cv::Point(80, 0));

Use radius_nearest().

vector<node<cv::Point> *> neighbors = tree->radius_nearest(cv::Point(70, 0), 100);

In this example the result is {(80, 0), (40, 0), (20, 0), (10, 0), (160, 0)}.

Use k_nearest().

vector<node<cv::Point> *> neighbors = tree->k_nearest(cv::Point(70, 0), 5);

In this example the result is {(80, 0), (40, 0), (20, 0), (10, 0), (160, 0)}.

main.cpp includes primitive unit tests with assert.
You can run the test by using make clean run.

kdtree is released under the MIT License, see LICENSE.txt.