3D plane detection for robot perception applying particle swarm optimization

Abstract

This article describes a 3D plane detection method for an intelligent robot to perceive an unknown object with 3D range sensor. Previously, various method has been proposed to perceive unknown environment. However, the previous unknown object detection method has problems which are high computational costs and low-accuracy for small object. In order to solve the mentioned problems, we have proposed an online processable unknown object detection based on a 3D plane detection method. The proposed method consists of simple plane detection applying particle swarm optimization (PSO) with region growing (RG) and integrated object plane detection. The simple plane detection is focused on small plane detection and reducing computational costs. To improve the accuracy, we apply PSO and RG. And, integrated object plane detection focuses on stability of detecting plane. As experimental results, we show that the computational cost is reduced to be able to calculate in real time for robot operation. And, the proposed method detects small planes of specific objects. Furthermore, we discuss the capability of proposed method which coordinate the ability of reducing computational costs and improving the plane detection accuracy.