Iterative Closest Point (ICP)

Synonyms

ICP

Definition

Iterative closest point (ICP) is a popular algorithm employed to register two sets of curves, two sets of surfaces, or two clouds of points.

Background

The ICP technique was proposed independently by Besl and McKay [1] and Zhang [2] in two different contexts. Besl and McKay [1] developed the ICP algorithm to register partially sensed data from rigid objects with an ideal geometric model, prior to shape inspection. So this is a subset–set matching problem because each sensed point has a correspondence in the ideal model. Zhang [2] developed the ICP algorithm in the context of autonomous vehicle navigation in rugged terrain based on vision. His algorithm is used to register a sequence of sensed data in order to build a complete model of the scene and to plan a free path for navigation. So this is a subset–subset matching problem because a fraction of data in one set does not have any correspondence in the other set. To address this issue, Zhang’s ICP algorithm...