Model globally, match locally: Efficient and robust 3D object recognition

By Bertram Drost, Markus Ulrich, Nassir Navab, Slobodan Ilic. 2010

1,085 citations recorded by the source corpus.

This paper addresses the problem of recognizing free-form 3D objects in point clouds. Compared to traditional approaches based on point descriptors, which depend on local information around points, we propose a novel method that creates a global model description based on oriented point pair features and matches that model locally using a fast voting scheme. The global model description consists of all model point pair features and represents a mapping from the point pair feature space to the model, where similar features on the model are grouped together. Such representation allows using much sparser object and scene point clouds, resulting in very fast performance. Recognition is done locally using an efficient voting scheme on a reduced two-dimensional search space. We demonstrate the efficiency of our approach and show its high recognition performance in the case of noise, clutter and partial occlusions. Compared to state of the art approaches we achieve better recognition rates, and demonstrate that with a slight or even no sacrifice of the recognition performance our method is much faster then the current state of the art approaches.

Subfields: Aerospace Engineering, Environmental Engineering, Geology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

In Environmental Engineering

In Geology

Citations in the corpus

BOP: Benchmark for 6D Object Pose Estimation (2018, methodology): “Inspired by improvements of [15], the matching strategy of [9] was improved by mitigating the effect of the feature discretization step.”
A Survey of Surface Reconstruction from Point Clouds (2017, methodology): “We note that although this method can detect a small set of easily parame-terizable shapes, efﬁcient pose detection methods for arbitrary shapes can also be used [DUNI10].”
DenseFusion: 6D Object Pose Estimation by Iterative Dense Fusion (2019, methodology): “It is widely adopted by both classical methods [4, 8, 37] and recent learning-based approaches [17, 31, 34].”
Deep Symmetry Networks (2014, background): “There are a number of algorithms that solve for a maximal feature location in symmetry space but they are not efficient when the feature weights are frequently adjusted [9, 19].”
Voting-based pose estimation for robotic assembly using a 3D sensor (2012, methodology): “As shown in [23], we globally model an object using a set of all possible pair features computed from the object model.”
3D Object Recognition in Cluttered Scenes with Local Surface Features: A Survey (2014, positive): “technique include [4], [9], [44], [46], [70] and [127].”
Global Hypothesis Generation for 6D Object Pose Estimation (2016, methodology): “We will omit the methods [3, 19, 5, 33, 11] since they were already discussed in the”
Detection-based object labeling in 3D scenes (2012, background): “An alternative approach is to match existing 3D object models into a scene [9], [12], [8], [14].”
SLAM++: Simultaneous Localisation and Mapping at the Level of Objects (2013, methodology): “After peak ﬁnding, pose estimates for each scene reference point are clustered on the CPU according to translation and rotation thresholds as in [6].”
Latent-Class Hough Forests for 3D Object Detection and Pose Estimation (2014, methodology): “State-of-the-art approaches in 3D detection and pose estimation [7,16] avoid this issue by training just from 3D models of the target object.”
BOP Challenge 2020 on 6D Object Localization (2020, methodology): “87 17 Drost-CVPR10-3D-Only-Faster [10] 45.”
Robotic roommates making pancakes (2011, neutral): “[15].”
Discriminatively Trained Templates for 3D Object Detection: A Real Time Scalable Approach (2013, neutral): “Approach DTT-3D Hinterstoisser[7] Drost [3]”
Learning 6D Object Pose Estimation Using 3D Object Coordinates (2014, background): “[4] also take a voting approach, and use oriented point pair features to reduce the search space.”
3DMatch: Learning Local Geometric Descriptors from RGB-D Reconstructions (2016, neutral): “[10] 5.”