Filtering Organized 3D Point Clouds for Bin Picking Applications

Published

January 23, 2024

Author(s)

Marek Franaszek, Prem Rachakonda, Kamel S. Saidi

Abstract

In many manufacturing applications, autonomous robot action is guided by a perception system integrated with the robot. The quality of the acquired 3D data must be high to ensure smooth and uninterrupted execution of robotic tasks. Unfortunately, many perception systems, mostly based on laser scanning or structured light technologies, output data contaminated by spurious points, which have no correspondence to the real physical objects. In bin picking applications, spurious points in 3D data are the outliers that may spoil obstacle avoidance planning executed by the robot controller and impede segmentation of individual parts in the bin. This may lead to failed attempts of gripping these parts. Many outlier removal procedures have been proposed which work very well on 3D point clouds acquired for different, mostly outdoor, scenarios, but these usually do not transfer well to the manufacturing domain. This paper presents a new filtering technique specifically designed to deal with a cluttered scene, which is typical for a bin picking task.

Citation

Applied Sciences

Volume

14

Pub Type

Journals

Download Paper

https://doi.org/10.3390/app14030961

Local Download

Keywords

bin picking, 3D point cloud, vision system and robot, outlier removal, segmentation

Citation

Franaszek, M. , Rachakonda, P. and Saidi, K. (2024), Filtering Organized 3D Point Clouds for Bin Picking Applications, Applied Sciences, [online], https://doi.org/10.3390/app14030961, https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=956513 (Accessed December 11, 2024)