ObjectMatch: Robust Registration using Canonical Object Correspondences

12/05/2022
by   Can Gümeli, et al.
0

We present ObjectMatch, a semantic and object-centric camera pose estimation for RGB-D SLAM pipelines. Modern camera pose estimators rely on direct correspondences of overlapping regions between frames; however, they cannot align camera frames with little or no overlap. In this work, we propose to leverage indirect correspondences obtained via semantic object identification. For instance, when an object is seen from the front in one frame and from the back in another frame, we can provide additional pose constraints through canonical object correspondences. We first propose a neural network to predict such correspondences on a per-pixel level, which we then combine in our energy formulation with state-of-the-art keypoint matching solved with a joint Gauss-Newton optimization. In a pairwise setting, our method improves registration recall of state-of-the-art feature matching from 77 overall and from 21 registering RGB-D sequences, our method outperforms cutting-edge SLAM baselines in challenging, low frame-rate scenarios, achieving more than 35 trajectory error in multiple scenes.

READ FULL TEXT

page 1

page 3

page 6

page 8

page 11

page 12

research
07/16/2022

Level Set-Based Camera Pose Estimation From Multiple 2D/3D Ellipse-Ellipsoid Correspondences

In this paper, we propose an object-based camera pose estimation from a ...
research
03/29/2019

FrameNet: Learning Local Canonical Frames of 3D Surfaces from a Single RGB Image

In this work, we introduce the novel problem of identifying dense canoni...
research
06/28/2022

3D Multi-Object Tracking with Differentiable Pose Estimation

We propose a novel approach for joint 3D multi-object tracking and recon...
research
03/04/2017

Automated Top View Registration of Broadcast Football Videos

In this paper, we propose a novel method to register football broadcast ...
research
02/05/2019

A Practical Maximum Clique Algorithm for Matching with Pairwise Constraints

A popular paradigm for 3D point cloud registration is by extracting 3D k...
research
04/26/2019

GN-Net: The Gauss-Newton Loss for Deep Direct SLAM

Direct methods for SLAM have shown exceptional performance on odometry t...
research
03/26/2019

High-quality Instance-aware Semantic 3D Map Using RGB-D Camera

We present a mapping system capable of constructing detailed instance-le...

Please sign up or login with your details

Forgot password? Click here to reset