A Javascript library written in Typescript to perform the iterative closest point (ICP) algorithm on two 2D point clouds.
npm install @eriknoorland/icpjs
import icpjs from '@eriknoorland/icpjs';
const reference = [{ x, y }];
const points = [{ x, y }];
const result = icpjs.run(reference, points);
console.log(result);The reference is a 2D point cloud or line segments array that is used as a map for the algorithm to find the translation and rotation it needs to perform on the points data.
These are the points you would like to match against the reference and find the translation and rotation for to align the points cloud to the reference cloud.
An object { x, y, phi } representing the estimated pose of where the data points where taken from. It defaults to { x: 0, y: 0, phi: 0 }.
This is used to set the maximum number of times the algorithm will try to find the best possible transformation and rotation. The default value is 20.
This is the amount the algorithm is allowed to be off to still be considered successful. The default value is 0.1.
An object containing the transformation object, an array of transformed data points and the number of itereations it took to get to the result.