A header-only implementation of an AABB tree.
Collision detection is common in many applications, especially games. Subsequently, collision detection often require checking for potential collision between many different game objects, and as such it is important that the collision detection is efficient. A naive implementation will end up with quadratic complexity, which isn't very scalable. By using an AABB tree, bounding hitboxes are stored in a tree according to their position and size, which in turn enables logarithmic complexity for collision detection.
#include <abby.hpp>
#include <iterator> // back_inserter
#include <vector> // vector
#include <iostream> // clog
void foo()
{
// Constructs an AABB tree that uses integers as identifiers and AABBs with double precision
abby::tree<int, double> tree;
// Inserts a few AABBs, takes a lower and upper bound
tree.insert(1, {150.0, 165.0}, {50.0, 50.0});
tree.insert(2, {88.0, 63.0}, {150.0, 343.0});
tree.insert(3, {412.0, 132.0}, {566.0, 291.0});
// Could also use a stack buffer
std::vector<int> candidates;
// Find collision candidates
tree.query(1, std::back_inserter(candidates));
for (const auto candidate : candidates) {
// Obtains an AABB
const auto& aabb = tree.get_aabb(candidate);
}
// Prints the tree using a stream
tree.print(std::clog);
// Updates an existing AABB
tree.update(2, {33.0, 76.0}, {123.0, 155.0});
// Sets the position of an AABB
tree.relocate(2, {12.0, 34.0});
// Removes an AABB from the tree
tree.erase(2);
// Rebuild as optimal tree
tree.rebuild();
// Clear the tree of all entries
tree.clear();
}
This library is an adapted and improved version of the AABBCC library. Furthermore, the AABB tree in the Simple Voxel Engine project also influenced the design of this library.