Robust real-time visual odometry for dense RGB-D mapping

By Thomas Whelan, Hordur Johannsson, Michael Kaess, John J. Leonard, John McDonald Jr.. 2013

385 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition, 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 Computer Vision and Pattern Recognition

In Geology

Citations in the corpus

Deep Sliding Shapes for Amodal 3D Object Detection in RGB-D Images (2015, background): “Note that we can also encode colors in this 3D volumetric representation, by appending RGB values to each voxel [28].”
Deep learning for detecting robotic grasps (2013, background): “[66] have extended and improved Simultaneous Localization and Mapping”
A benchmark for RGB-D visual odometry, 3D reconstruction and SLAM (2014, methodology): “We evaluate 5 different odometers; (i) DVO [13] which performs full frame alignment between pairs of RGB-D frames making use of a robust cost function to improve robustness to outliers; (ii) FOVIS [14] which uses FAST feature correspondences between pairs of RGB-D frames to resolve camera motion, re”
Large-Scale Multi-resolution Surface Reconstruction from RGB-D Sequences (2013, methodology): “This stands in contrast to [18, 19] where the terms are additively combined using a heuristically chosen weight.”
Chisel: Real Time Large Scale 3D Reconstruction Onboard a Mobile Device using Spatially Hashed Signed Distance Fields (2015, methodology): “Some works have either used octrees [24] [30] [31], or use a moving volume[12].”
Visual-lidar odometry and mapping: low-drift, robust, and fast (2015, background): “Motion estimation with RGB-D cameras [8], [9] can be conducted easily with scale.”
3-D Mapping With an RGB-D Camera (2014, background): “Kintinuous [21] overcomes this limitation by virtually moving the voxel grid with the current camera pose.”
Online Reconstruction of Indoor Scenes from RGB-D Streams (2016, methodology): “Robotics researchers in general formulate the problem in a SLAM framework [8, 34, 12, 13, 31, 32], minimizing the absolute trajectory error (ATE) is their main focus and the root mean square error (RMSE) of the ATE is used as the evaluation metric.”
Volumetric 3D mapping in real-time on a CPU (2014, methodology): “To overcome the problem of memory limitation, several authors proposed shifting reconstruction volumes [15], [18], where older parts of the SDF are converted into a triangle mesh using Marching Cubes [13].”
Elastic Fragments for Dense Scene Reconstruction (2013, background): “These ideas have subsequently been extended to larger environments [34, 35].”
Robust reconstruction of indoor scenes (2015, methodology): “We evaluated models reconstructed by Kintinuous, DVO SLAM, SUN3D SfM, and our approach.”
Real-time large-scale dense RGB-D SLAM with volumetric fusion (2014, methodology): “Note that the results on the fr2/large no loop dataset are taken from our previous work, Whelan et al. (2013a).”
REMODE: Probabilistic, monocular dense reconstruction in real time (2014, background): “In robotics, the use of RGBD cameras is favouring the development of techniques for highly-detailed [7] and spatially-extended reconstructions [8], their applicability being limited to short range measurements and indoor environments.”
An Octree-Based Approach towards Efficient Variational Range Data Fusion (2016, methodology): “To compute the coloring of the resulting mesh, we weigh for each vertex the reprojected colors according to the dot product between its normal and every camera view vector, similar to [24].”
Dense visual SLAM for RGB-D cameras (2013, methodology): “Although this problem can be delayed using more elaborate cost functions [7], [8] and alignment procedures [9], a more fundamental solution is desirable.”
SceneNet: An annotated model generator for indoor scene understanding (2016, methodology): “Such scenes may be valuable for benchmarking the large scale SLAM systems that have been developed in recent years [2], [29].”