2D Object Localization Based Point Pair Feature for Pose Estimation

Liu, Diyi; Arai, Shigeo; Zhao, Feng; Miao, Jiaqi; Xu, Yajun; Kinugawa, Jun; Kosuge, Kazuhiro

doi:10.1109/robio.2018.8665097

Cited by 19 publications

(12 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Hinterstoisser et al [25] introduced a better and more efficient sampling strategy with modifications to the pre and post-processing steps, which achieved good results. Liu et al obtained an impressive result by combining a machine learning-based 2D object localization and a 3D pose estimation method in [26], [27]. Vidal et al proposed a better preprocessing and clustering step with an improved matching method for Point Pair Features in [29].…”

Section: A Feature-based Methodsmentioning

confidence: 99%

Fast Object Pose Estimation Using Adaptive Threshold for Bin-Picking

Yan

Zhou

et al. 2020

IEEE Access

View full text Add to dashboard Cite

Robotic bin-picking is a common process in modern manufacturing, logistics, and warehousing that aims to pickup known or unknown objects with random poses out of a bin by using a robot-camera system. Rapid and accurate object pose estimation pipelines have become an escalating issue for robot picking in recent years. In this paper, a fast 6-DoF (degrees of freedom) pose estimation pipeline for random bin-picking is proposed in which the pipeline is capable of recognizing different types of objects in various cluttered scenarios and uses an adaptive threshold segment strategy to accelerate estimation and matching for the robot picking task. Particularly, our proposed method can be effectively trained with fewer samples by introducing the geometric properties of objects such as contour, normal distribution, and curvature. An experimental setup is designed with a Kinova 6-Dof robot and an Ensenso industrial 3D camera for evaluating our proposed methods with respect to four different objects. The results indicate that our proposed method achieves a 91.25% average success rate and a 0.265s average estimation time, which sufficiently demonstrates that our approach provides competitive results for fast objects pose estimation and can be applied to robotic random bin-picking tasks. INDEX TERMS Robotic bin-picking, learning-based pose estimation, adaptive threshold, point pair features, multilayer segmentation, 3D sensing.

show abstract

Section: A Feature-based Methodsmentioning

confidence: 99%

Fast Object Pose Estimation Using Adaptive Threshold for Bin-Picking

Yan

Zhou

et al. 2020

IEEE Access

View full text Add to dashboard Cite

show abstract

“…PBVS is a method of positioning a robot through the minimization of the difference between target and current poses of the robot which is estimated from captured images. PBVS has been attracting attention due to the recent price reduction and spread of 3D sensors, and the progress of 3D measurement [15], [16] and pose estimation [17], [18], [19] technology. However, PBVS requires an intrinsic parameter, which results in the vulnerability to the errors of the camera parameters.…”

Section: Introductionmentioning

confidence: 99%

Convolutional Neural Network based Visual Servoing for Eye-to-Hand Manipulator

Arai¹,

Kosuge²

2022

Preprint

Self Cite

View full text Add to dashboard Cite

We propose a CNN based visual servoing scheme for precise positioning of an eye-to-hand manipulator in which the control input of a robot is calculated directly from images by a neural network. In this paper, we propose Difference of Encoded Features driven Interaction matrix Network (DEFINet), a new convolutional neural network (CNN), for eye-to-hand visual servoing. DEFINet estimates a relative pose between desired and current end-effector from desired and current images captured by an eye-to-hand camera. DEFINet includes two branches of the same CNN that share weights and encode target and current images, which is inspired by the architecture of Siamese network. Regression of the relative pose from the difference of the encoded target and current image features leads to a high positioning accuracy of visual servoing using DEFINet. The training dataset is generated from sample data collected by operating a manipulator randomly in task space. The performance of the proposed visual servoing is evaluated through numerical simulation and experiments using a six-DOF industrial manipulator in a real environment. Both simulation and experimental results show the effectiveness of the proposed method.<br>

show abstract

“…For example, Liu et al [11] locate the object in the image using convolutional neural networks (CNN) [14]- [17], and then obtain the corresponding point cloud through bounding box or mask [9]- [11]. Li et al [1] and Li and Hashimoto [13] simply divided the point cloud into many regions of interest (ROI).…”

Section: Introductionmentioning

confidence: 99%

A Convolutional Neural Network for Point Cloud Instance Segmentation in Cluttered Scene Trained by Synthetic Data Without Color

Arai

Kosuge

2020

IEEE Access

Self Cite

View full text Add to dashboard Cite

3D Instance segmentation is a fundamental task in computer vision. Effective segmentation plays an important role in robotic tasks, augmented reality, autonomous driving, etc. With the ascendancy of convolutional neural networks in 2D image processing, the use of deep learning methods to segment 3D point clouds receives much attention. A great convergence of training loss often requires a large amount of human-annotated data, while making such a 3D dataset is time-consuming. This paper proposes a method for training convolutional neural networks to predict instance segmentation results using synthetic data. The proposed method is based on the SGPN framework. We replaced the original feature extractor with ''dynamic graph convolutional neural networks'' that learned how to extract local geometric features and proposed a simple and effective loss function, making the network more focused on hard examples. We experimentally proved that the proposed method significantly outperforms the state-of-the-art method in both Stanford 3D Indoor Semantics Dataset and our datasets.

show abstract

2D Object Localization Based Point Pair Feature for Pose Estimation

Cited by 19 publications

References 17 publications

Fast Object Pose Estimation Using Adaptive Threshold for Bin-Picking

Fast Object Pose Estimation Using Adaptive Threshold for Bin-Picking

Convolutional Neural Network based Visual Servoing for Eye-to-Hand Manipulator

A Convolutional Neural Network for Point Cloud Instance Segmentation in Cluttered Scene Trained by Synthetic Data Without Color

Contact Info

Product

Resources

About