Dense visual SLAM for RGB-D cameras

By Christian Kerl, Jürgen Sturm, Daniel Cremers. 2013 IEEE/RSJ International Conference on Intelligent Robots and Systems, 2013

985 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/4b838aa7b06b202242916ecaf1387d92e52a40df

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

Citations in the corpus

ORB-SLAM2: An Open-Source SLAM System for Monocular, Stereo, and RGB-D Cameras (2016, methodology): “In Table III we compare our accuracy to the following state-of-the-art methods: ElasticFusion [15], Kintinuous [12], DVO-SLAM [14] and RGB-D SLAM [13].”
LDSO: Direct Sparse Odometry with Loop Closure (2018, methodology): “Thus, direct VO systems have in the past been extended to SLAM either by using only keyframe proximity for loop closure detection [6] or by computing features for loop closure detection independently from frontend tracking and constraint computation [7].”
A Comprehensive Survey of Visual SLAM Algorithms (2022, methodology): “The dense visual odometry SLAM (DVO-SLAM) algorithm, proposed by Kerl et al. [95], is a keyframe-based technique.”
Back to the Feature: Learning Robust Camera Localization from Pixels to Pose (2021, background): “Like their classical counterparts [26,38], they also suffer from a small basin of convergence, limiting them to frame tracking.”
EM-Fusion: Dynamic Object-Level SLAM With Probabilistic Data Association (2019, methodology): “alignment [4], and map representations such as surfels [9] or keyframes [10].”
DeepV2D: Video to Depth with Differentiable Structure from Motion (2018, background): “360 desk desk2 plant room rpy xyz mean DVO (Kerl et al., 2013) 0.”
FAST-LIO2: Fast Direct LiDAR-Inertial Odometry (2021, background): “On the other hand, real-time dense mapping [5]–[8] based on visual sensors at high resolution and accuracy with only the robot onboard computation resources is still a grand challenge.”
ElasticFusion: Dense SLAM Without A Pose Graph (2015, methodology): “System kt0 kt1 kt2 kt3 DVO SLAM 0.104m 0.029m 0.191m 0.152m RGB-D SLAM 0.026m 0.008m 0.018m 0.433m MRSMap 0.204m 0.228m 0.189m 1.090m Kintinuous 0.072m 0.005m 0.010m 0.355m Frame-to-model 0.497m 0.009m 0.020m 0.”
RTAB‐Map as an open‐source lidar and visual simultaneous localization and mapping library for large‐scale and long‐term online operation (2018, positive): “Table 6 presents ATE results with additional ones from other approaches like ElasticFusion (Whelan et al., 2016), Kintin-uous (Whelan et al., 2015), DVO ‐ SLAM (Kerl et al., 2013), RGBDSLAMv2 (Endres et al., 2014),”
Large-scale direct SLAM with stereo cameras (2015, background): “In the RGB-D domain [8], [9], [10], direct methods have become the state-of-the-art for their high accuracy and efficiency.”
School of Electrical and Computer Engineering Oklahoma State University , Stillwater , OK 74078 , USA (methodology): “Our method is also different from Kintinuous [24], DVO SLAM [17], SUN3D SfM [26], and RGB-D mapping [14], which involve visual odometry and use frameto-frame tracking and matching system in their pipelines.”
Robust reconstruction of indoor scenes (2015, methodology): “(For these experiments, all settings for imagebased loop closure [34] were set to maximize accuracy, and loop closure detection was performed between every single pair of frames.”
Dense Contrastive Learning for Self-Supervised Visual Pre-Training (2020, background): “The visual correspondence problem is to compute the pairs of pixels from two images that result from the same scene [43], and it is crucial for many applications, including optical flow [8], structure-frommotion [35], visual SLAM [20], 3D reconstruction [10] etc.”
A General Optimization-based Framework for Local Odometry Estimation with Multiple Sensors (2019, methodology): “We have seen many impressive approaches that work with one kind of sensor, such as visual-based methods [1]–[5], LiDAR-based methods [13], RGB-D based methods [14]. and event-based methods [15].”
PL-SLAM: A Stereo SLAM System Through the Combination of Points and Line Segments (2017, methodology): “For deciding when a new KF is inserted in the map, we have followed the approach in [45], which employs the uncertainty”
Unmanned Aerial Vehicles for Search and Rescue: A Survey (2023, positive): “The paper [141] suggests the utilization of RGB-D cameras to obtain dense color and depth images [142] for an onboard UAV SLAM approach.”
SfM-Net: Learning of Structure and Motion from Video (2017, methodology): “Similarly, direct meth-ods for visual SLAM having RGB [26] or RGBD [17] video as input, perform dense pixel matching from frame to frame while estimating a camera trajectory and depth of the pixels in the scene with impressive 3D point cloud reconstructions.”
D3VO: Deep Depth, Deep Pose and Deep Uncertainty for Monocular Visual Odometry (2020, methodology): “…and the direct VO share a similar photometric objective, we propose to use the learned weights to replace the weighting function of the photometric residual in traditional direct VO which is empirically set [61] or only accounts for the intrinsic uncertainty of the speciﬁc algorithm itself [16,37].”
Vox-Fusion: Dense Tracking and Mapping with Voxel-based Neural Implicit Representation (2022, methodology): “Following this basic design, many systems gain improvements by introducing different 3D structures [38], exploring space subdivision [21, 24] and performing global map optimization [11,13,26].”