Optimization of Registration Performance Metrics

Published

February 25, 2016

Author(s)

Marek Franaszek, Geraldine Cheok

Abstract

Perception systems are a very important component of automation and robotics. These systems localize objects in its local coordinate frame and the data has to be available for use in another coordinate frame. This requires that the data be transformed from one coordinate frame to another and the transformation has to be determined. There are different methods to obtain this transformation. In this study, rigid body, point-based registration is investigated. This report details the effort to develop guidelines for this type of registration based on physical experiments and noise characterization. This work also involves the development of performance metrics to quantify the quality of the registration and guidance to select the location and number of points for optimal registration.

Citation

NIST Interagency/Internal Report (NISTIR) - 8111

Report Number

8111

NIST Pub Series

NIST Interagency/Internal Report (NISTIR)

Pub Type

NIST Pubs

Download Paper

DOI Link

Keywords

fiducials, performance metrics, point-based registration, registration error, uncertainty

Citation

Franaszek, M. and Cheok, G. (2016), Optimization of Registration Performance Metrics, NIST Interagency/Internal Report (NISTIR), National Institute of Standards and Technology, Gaithersburg, MD, [online], https://doi.org/10.6028/NIST.IR.8111 (Accessed April 19, 2024)