Paper Info
Title
3D Object Discovery and Modeling Using Single RGB-D Images Containing Multiple Object Instances
Abstract
Unsupervised object modeling is important in robotics, especially for handling a large set of objects. We present a method for unsupervised 3D object discovery, reconstruction, and localization that exploits multiple instances of an identical object contained in a single RGB-D image. The proposed method does not rely on segmentation, scene knowledge, or user input, and thus is easily scalable. Our method aims to find recurrent patterns in a single RGB-D image by utilizing appearance and geometry of the salient regions. We extract keypoints and match them in pairs based on their descriptors. We then generate triplets of the keypoints matching with each other using several geometric criteria to minimize false matches. The relative poses of the matched triplets are computed and clustered to discover sets of triplet pairs with similar relative poses. Triplets belonging to the same set are likely to belong to the same object and are used to construct an initial object model. Detection of remaining instances with the initial object model using RANSAC allows to further expand and refine the model. The automatically generated object models are both compact and descriptive. We show quantitative and qualitative results on RGB-D images with various objects including some from the Amazon Picking Challenge. We also demonstrate the use of our method in an object picking scenario with a robotic arm.
Year
DOI
Venue
2017
10.1109/3DV.2017.00056
2017 International Conference on 3D Vision (3DV)
Keywords
DocType
Volume
Computer-vision,Robotics,RGB-D,Unsupervised,Discovery,Keypoints,Matching,Reconstruction,Pose-estimation
Conference
abs/1710.06231
ISSN
ISBN
Citations 
2378-3826
978-1-5386-2611-5
2
PageRank 
References 
Authors
0.44
20
5
Name
Order
Citations
PageRank
Wim Abbeloos142.84
Esra Ataer Cansizoglu2193.58
Sergio Caccamo362.88
Yuichi Taguchi451839.71
Domae, Y.5223.36