A fast 3D object recognition algorithm using plane-constrained point pair features

  • PDF / 4,745,945 Bytes
  • 21 Pages / 439.37 x 666.142 pts Page_size
  • 102 Downloads / 207 Views

DOWNLOAD

REPORT


A fast 3D object recognition algorithm using plane-constrained point pair features Zhengtao Xiao 1,2

1

3

1

& Jian Gao & Dongqing Wu & Lanyu Zhang & Xin Chen

1

Received: 21 October 2019 / Revised: 27 July 2020 / Accepted: 4 August 2020 # Springer Science+Business Media, LLC, part of Springer Nature 2020

Abstract

The point pair feature (PPF) algorithm is one of the best-performing 3D object recognition algorithms. However, the high dimensionality of its search space is a disadvantage of this algorithm. This high dimensionality means the feature matching process contains a large number of uninformative features, which reduces recognition speed. To solve this problem and improve the object recognition speed, this paper proposes a fast 3D object recognition algorithm based on the plane-constrained point pair features. By utilizing the property of the coplanar point pair features and the characteristics of the object placement plane, the proposed algorithm extracts the object placement plane through convex hull area calculation, eliminates irrelevant point pair features, and then performs object recognition with the reduced point pair feature descriptors for the feature matching. Experimental results demonstrate that the proposed algorithm significantly reduces the number of feature descriptors and accelerates the recognition speed of 3D objects in a complex background. Compared to the original point pair feature algorithm, the proposed method can achieve better performance and efficiency for 3D object recognition. Keywords 3D object recognition . Point pair features . Feature descriptor . Feature matching . Convex hull

* Jian Gao [email protected]

1

State Key Laboratory of Precision Electronic Manufacturing Technology and Equipment, Guangdong University of Technology, Guangzhou 510006, China

2

School of Electromechanical Engineering, Guangdong Polytechnic of Industry and Commerce, Guangzhou 510510, China

3

School of Computational Science, Zhongkai University of Agriculture and Engineering, Guangzhou 510220, China

Multimedia Tools and Applications

1 Introduction In recent years, based on the development and popularization of depth sensing technology, the acquisition of depth images has become significantly easier. Additionally, 3D object recognition technology based on depth images has also received extensive attention and seen various applications [35]. Currently, 3D object recognition in complex scenes is one of the most important tasks in computer vision and it is widely used in robotics, unmanned vehicles, security, etc. Unlike traditional 2D object recognition methods, 3D object recognition can obtain distance information from depth images [21] or point clouds [31]. Feature extraction is robust to scale, rotation, or illumination [3, 17]. Furthermore, distance information that cannot be acquired from RGB images can be obtained directly from depth images or point clouds. Based on these advantages, studies on 3D object recognition have become a hot topic in recent years. The goal of 3D object recog