DTAM: Dense tracking and mapping in real-time
The DOI matches its Crossref record; 16 of the 17 references it cites were found in a registry or an index.
Abstract
DTAM is a system for real-time camera tracking and reconstruction which relies not on feature extraction but dense, every pixel methods. As a single hand-held RGB camera flies over a static scene, we estimate detailed textured depth maps at selected keyframes to produce a surface patchwork with millions of vertices. We use the hundreds of images available in a video stream to improve the quality of a simple photometric data term, and minimise a global spatially regularised energy functional in a novel non-convex optimisation framework. Interleaved, we track the camera's 6DOF motion precisely by frame-rate whole image alignment against the entire dense model. Our algorithms are highly parallelisable throughout and DTAM achieves real-time performance using current commodity GPU hardware. We demonstrate that a dense model permits superior tracking performance under rapid motion compared to a state of the art method using features; and also show the additional usefulness of the dense model for real-time scene interaction in a physics-enhanced augmented reality application.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“However, in our network, multi-scale projections not only help to overcome gradient locality during training (see [33], [47] for a broader discussion of this well-known issue), but also aid in online error correction at runtime.”
From Unsupervised Deep Visual-Inertial Odometry with Online Error Correction for RGB-D Imagery (2020) · cites this paper as [33]
In the citing paper
“SFMLearner [33] demonstrated the unsupervised learning of unscaled egomotion and depth from RGB imagery.”
From Unsupervised Deep Visual-Inertial Odometry with Online Error Correction for RGB-D Imagery (2020) · cites this paper as [33]
In the citing paper
“For KITTI, VIOLearner RGB outperforms approaches SFMLearner [33] and the network of Zhan et al. [30].”
From Unsupervised Deep Visual-Inertial Odometry with Online Error Correction for RGB-D Imagery (2020) · cites this paper as [33]
In the citing paper
“For AirSim, VIOLearner (no IMU) outperforms approaches SFMLearner [33] and ORB-SLAM2-S in translation error.”
From Unsupervised Deep Visual-Inertial Odometry with Online Error Correction for RGB-D Imagery (2020) · cites this paper as [33]
In the citing paper
“Compared to other vision-only approaches, for KITTI, VIOLearner (no IMU) outperforms approaches SFMLearner [33] and the network of Zhan et al. [30] and yet is outperformed by ORB-SLAM-2 which uses full bundle adjustment.”
From Unsupervised Deep Visual-Inertial Odometry with Online Error Correction for RGB-D Imagery (2020) · cites this paper as [33]
Citations
Cites
Nonlinear total variation based noise removal algorithms
Rudin, L. I., Osher, S., & Fatemi, E. (1992) · Physica D: Nonlinear Phenomena · doi:10.1016/0167-2789(92)90242-f · Cited by 15,704
Parallel Tracking and Mapping for Small AR Workspaces
Klein, G., & Murray, D. (2007) · 2007 6th IEEE and ACM International Symposium on Mixed and Augmented Reality · doi:10.1109/ismar.2007.4538852 · Cited by 4,346
A First-Order Primal-Dual Algorithm for Convex Problems with Applications to Imaging
Chambolle, A., & Pock, T. (2011) · Journal of Mathematical Imaging and Vision · doi:10.1007/s10851-010-0251-1 · Cited by 4,336
Lucas-Kanade 20 Years On: A Unifying Framework
Baker, S., & Matthews, I. (2004) · International Journal of Computer Vision · doi:10.1023/b:visi.0000011205.11775.fd · Cited by 3,155
Fast cost-volume filtering for visual correspondence and beyond
Rhemann, C., Hosni, A., Bleyer, M., Rother, C., & Gelautz, M. (2011) · CVPR 2011 · doi:10.1109/cvpr.2011.5995372 · Cited by 485
Live dense reconstruction with a single moving camera
Newcombe, R. A., & Davison, A. J. (2010) · 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2010.5539794 · Cited by 366
Self-citationImproving the Agility of Keyframe-Based SLAM
Klein, G., & Murray, D. (2008) · Lecture Notes in Computer Science · doi:10.1007/978-3-540-88688-4_59 · Cited by 335
Robust multi-sensor image alignment
Irani, M., & Anandan, P. (2002) · Sixth International Conference on Computer Vision (IEEE Cat. No.98CH36271) · doi:10.1109/iccv.1998.710832 · Cited by 220
Real-Time Dense Geometry from a Handheld Camera
Stühmer, J., Gumhold, S., & Cremers, D. (2010) · Lecture Notes in Computer Science · doi:10.1007/978-3-642-15986-2_2 · Cited by 205
Some First-Order Algorithms for Total Variation Based Image Restoration
Aujol, J. (2009) · Journal of Mathematical Imaging and Vision · doi:10.1007/s10851-009-0149-y · Cited by 151
Cited by
Items 801–900 of 1,004
Object-Centric Photometric Bundle Adjustment with Deep Shape Prior
Zhu, R., Wang, C., Lin, C.-H., Wang, Z., & Lucey, S. (2018) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv.2018.00103 · Cited by 15
Context‐Aware Mixed Reality: A Learning‐Based Framework for Semantic‐Level Interaction
Chen, L., Tang, W., John, N., Wan, T.-R., & Zhang, J. J. (2019) · Computer graphics forum (Print) · doi:10.1111/cgf.13887 · Cited by 17
“Recent advances in depth sensors, display technologies and SLAM software [NLD11] [NIH*11] [WJK*13] [NFS15] have opened up the potential of MR systems.”
BackgroundMethodInfluentialMarchesi, G., Eichhorn, C., Plecher, D. A., Itoh, Y., & Klinker, G. (2021) · ISPRS International Journal of Geo-Information · doi:10.3390/ijgi10110772 · Cited by 16
MDN-VO: Estimating Visual Odometry with Confidence
Kaygusuz, N., Méndez, O., & Bowden, R. (2021) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros51168.2021.9636827 · Cited by 15
“[21] proposed a dense direct approach, which estimates the motion by optimising image pixel intensities.”
BackgroundPICCOLO: Point Cloud-Centric Omnidirectional Localization
Kim, J., Choi, C., Jang, H., & Kim, Y. M. (2021) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.00330 · Cited by 20
“In contrast, conventional photometric loss widely used in computer vision [14, 12] evaluates the color difference at every pixel location [12, 32], thus is image-centric .”
BackgroundMethodInfluentialHigh Fidelity Immersive Virtual Reality
Gilson, S. J., & Glennerster, A. (2012) · InTech eBooks · doi:10.5772/50655 · Cited by 15
“Results in this area even for cameras with very rapid, jerky movements, are impressive [19].”
Monocular vision-based autonomous navigation system on a toy quadcopter in unknown environments
Huang, R., Tan, P., & Chen, B. M. (2015) · International Conference on Unmanned Aircraft Systems (ICUAS) · doi:10.1109/icuas.2015.7152419 · Cited by 18
“[6] proposed a dense tracking and mapping system which can reconstruct dense surfaces of the scene using a single camera.”
BackgroundAugmented Reality Marker Hiding with Texture Deformation
Kawai, N., Sato, T., Nakashima, Y., & Yokoya, N. (2016) · IEEE Transactions on Visualization and Computer Graphics · doi:10.1109/tvcg.2016.2617325 · Cited by 15
“Although the SLAM-based approaches [2], [3], [4] have been intensively studied in these days, the marker-based approach [1] is widely and practically used because of its easiness of putting virtual objects on users’ desired positions and its robustness against various textures and shapes in the…”
BackgroundMethodStructure-aware SLAM with planes and lines in man-made environment
Zhang, J., Zeng, G., & Zha, H. (2018) · Pattern Recognition Letters · doi:10.1016/j.patrec.2018.10.037 · Cited by 14
FMD Stereo SLAM: Fusing MVG and Direct Formulation Towards Accurate and Fast Stereo SLAM
Tang, F., Li, H., & Wu, Y. (2019) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra.2019.8793664 · Cited by 17
“DTAM [28] is a system for real-time tracking and reconstruction by GPU which uses all pixels in the image.”
BackgroundDeep Probabilistic Feature-Metric Tracking
Xu, B., Davison, A. J., & Leutenegger, S. (2020) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2020.3039216 · Cited by 22
“D ENSE image alignment [1] using the photometric residual has been widely applied in 2D tracking [2], 3D object tracking [3], optical flow [4], and SLAM [5].”
InfluentialSelf-citationOnboard Real-Time Dense Reconstruction in Large Terrain Scene Using Embedded UAV Platform
Lai, Z., Liu, F., Guo, S., Meng, X., Han, S., & Li, W. (2021) · Remote Sensing · doi:10.3390/rs13142778 · Cited by 17
Robust Keyframe-based Dense SLAM with an RGB-D CameraPreprint
Liu, H., Li, C., Chen, G., Zhang, G., Kaess, M., & Bao, H. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1711.05166 · Cited by 20
“Although some methods [19]–[22] have been proposed to reconstruct dense 3D models online, it is still either limited to a small scene, or drift-free dense 3D reconstruction is not considered.”
MethodAn Overview of Perception and Decision-Making in Autonomous Systems in the Era of LearningPreprint
Tang, Y. S., Zhao, C., Wang, J., Zhang, C., Sun, Q., Zheng, W. X., Du, W., Qian, F., & Kurths, J. (2020) · arXiv (Cornell University) · Cited by 16
Rapid free-space mapping from a single omnidirectional camera
Lukierski, R., Leutenegger, S., & Davison, A. J. (2015) · European Conference on Mobile Robots · doi:10.1109/ecmr.2015.7324222 · Cited by 19
“With passive RGB cameras, recovering dense depth and free space information is much more challenging, but there has been good progress in the vast computer vision literature on multi-view stereo (MVS), and recently even in real-time with methods like DTAM [6].”
BackgroundMethodInfluentialSelf-citationCNN-MonoFusion: Online Monocular Dense Reconstruction Using Learned Depth from Single View
Wang, J., Liu, H., Cong, L., Xiahou, Z., & Wang, L. (2018) · IEEE International Symposium on Mixed and Augmented Reality Adjunct (ISMAR-Adjunct) · doi:10.1109/ismar-adjunct.2018.00034 · Cited by 15
“DPPTAM [5] adopts a piecewise planar assumption to fill the holes of surfaces recovered by DTAM [30].”
BackgroundEvaluation of 3D LiDAR SLAM algorithms based on the KITTI dataset
Wu, J., Huang, S., Yang, Y., & Zhang, B. (2023) · The Journal of Supercomputing · doi:10.1007/s11227-023-05267-3 · Cited by 14
Hu, Z., Yuan, J., Gao, Y., Wang, B., & Zhang, X. (2023) · IEEE Transactions on Intelligent Vehicles · doi:10.1109/tiv.2023.3303355 · Cited by 14
“To this end, some monocular dense SLAM works [12], [13], [14] proposed to integrate a depth prediction network into the VO or SLAM system.”
BackgroundLightedDepth: Video Depth Estimation in Light of Limited Inference View Angles
Zhu, S., & Liu, X. (2023) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52729.2023.00484 · Cited by 19
A Generalizable D-VIO and Its Fusion With GNSS/IMU for Improved Autonomous Vehicle Localization
Yusefi, A., Durdu, A., Bozkaya, F., Tığlıoğlu, Ş., Yılmaz, A., & Sungur, C. (2023) · IEEE Transactions on Intelligent Vehicles · doi:10.1109/tiv.2023.3316361 · Cited by 17
“Other meth-ods such as Dense Tracking and Mapping (DTAM) [19] measure each keyframe’s dense depth map to fit the picture to find the robot’s motion.”
BackgroundSemantic geometric fusion multi-object tracking and lidar odometry in dynamic environment
Ma, T., Jiang, G., Ou, Y., & Xu, S. (2024) · Robotica · doi:10.1017/s0263574723001868 · Cited by 16
“A 4D scene mapping module is also introduced for maintaining the long-term static map and movable object map separately, which can be used to further navigation applications.”
MethodGlobal-Context-Aware Visual Odometry System With Epipolar-Geometry-Constrained Loss Function
Ji, Z., Nan, K., Xu, Y., Wang, H., Li, X., & Bai, J. (2024) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2024.3370804 · Cited by 16
“Similar to [19], direct sparse odometry [20] is another direct method based on a fully direct probabilistic model using optimization algorithm and integrated photometric calibration to increase accuracy and robustness.”
MethodThree-Dimensional Dense Reconstruction: A Review of Algorithms and Datasets
Lee, Y. (2024) · Sensors · doi:10.3390/s24185861 · Cited by 20
“Dense Tracking And Mapping (DTAM) is a monocular SLAM method that generates dense depth maps for each frame in real time by minimizing the photometric error between the input image and a rendered version of the 3D model [37].”
MethodPhotorealistic 3D mapping of indoors by RGB-D scanning process
Tykkälä, T., Comport, A. I., & Kämäräinen, J. (2013) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros.2013.6696480 · Cited by 14
“DTAM was introduced as the dense version of PTAM, which allowed dense tracking and mapping using a monocular camera [5].”
BackgroundMethodWieringa, F. P., Bouma, H., Eendebak, P. T., van Basten, J. A., Beerlage, H. P., Smits, G. A. H. J., & Bos, J. E. (2014) · Journal of medical imaging · doi:10.1117/1.jmi.1.1.015001 · Cited by 14
“…to create a dense reconstruction from the input of video captured by a free moving handheld camera are already described in the literature; 11,12 however, it has not yet been applied to endoscopic surgery, which inherently suffers from additional problems, such as strong camera distortion…”
Video Registration to SfM Models
Kroeger, T., & Gool, L. V. (2014) · Lecture notes in computer science · doi:10.1007/978-3-319-10602-1_1 · Cited by 16
“However, it is an integral part of SLAM [8,19,30,29].”
BackgroundTamaazousti, M., Naudet-Collette, S., Gay‐Bellile, V., Bourgeois, S., Besbes, B., & Dhome, M. (2015) · Multimedia Tools and Applications · doi:10.1007/s11042-015-2968-8 · Cited by 15
“The first step relies on the KinectFusion process [35].”
MethodFeature article: an autonomous quadrotor avoiding a helicopter in low-altitude flights
Liu, Z., & Foina, A. G. (2016) · IEEE Aerospace and Electronic Systems Magazine · doi:10.1109/maes.2016.150131 · Cited by 16
Optimization Based Multi-View Coverage Path Planning for Autonomous Structure From Motion Recordings
Lindner, S., Garbe, C. S., & Mombaur, K. (2019) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2019.2926216 · Cited by 16
“[1]–[3]), they always suffer from a tradeoff between framerate, resolution and computational complexity.”
Active 3D Modeling via Online Multi-View Stereo
Song, S., Kim, D., & Jo, S. (2020) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra40945.2020.9197089 · Cited by 15
“The proposed online MVS system extends monocular mapping algorithms [11] [12] [13], which focus only on the local dense mapping, to be adaptable to constructing a large-scale model.”
BackgroundMethodA robust visual odometry based on RGB-D camera in dynamic indoor environments
Zhang, F., Li, Q., Wang, T., & Ma, T. (2020) · Measurement Science and Technology · doi:10.1088/1361-6501/abcc15 · Cited by 14
“Therefore, if the working environment is limited to a static, rigid body, illumination change that is not obvious and the environment has no human interference, the visual SLAM system is relatively mature, such as ORB SLAM2 [2], DTAM [3], PTAM [4] and so on.”
Deep Pose Graph-Matching-Based Loop Closure Detection for Semantic Visual SLAM
Duan, R., Feng, Y., & Wen, C. (2022) · Sustainability · doi:10.3390/su141911864 · Cited by 13
NARROW FIELD-OF-VIEW VISUAL ODOMETRY BASED ON A FOCUSED PLENOPTIC CAMERA
Zeller, N., Quint, F., & Stilla, U. (2015) · ISPRS annals of the photogrammetry, remote sensing and spatial information sciences · doi:10.5194/isprsannals-ii-3-w4-285-2015 · Cited by 16
“Different from other direct SLAM methods which used the complete image information for tracking and mapping (Newcombe et al., 2011), LSD-SLAM works only on image regions with sufficient contrast.”
BackgroundMethodReal-time camera tracking using a particle filter combined with unscented Kalman filters
Lee, S. (2014) · Journal of Electronic Imaging · doi:10.1117/1.jei.23.1.013029 · Cited by 14
A Joint 3D-2D Based Method for Free Space Detection on Roads
Patra, S., Maheshwari, P., Yadav, S., Banerjee, S., & Arora, C. (2018) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv.2018.00076 · Cited by 14
“These pairwise poses then can be used to estimate the structure and camera parameters simultaneously using SfM based Simultaneous Localization And Mapping (SLAM) framework [10, 22, 28, 29, 33].”
Scale-Awareness of Light Field Camera Based Visual Odometry
Zeller, N., Quint, F., & Stilla, U. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-01237-3_44 · Cited by 17
Online Spatial and Temporal Calibration for Monocular Direct Visual-Inertial Odometry
Feng, Z., Li, J., Zhang, L., & Chen, C. (2019) · Sensors · doi:10.3390/s19102273 · Cited by 16
“Some direct approaches such as DTAM [24] also initialize with a feature-based approach.”
MethodJoint optimization based on direct sparse stereo visual-inertial odometry
Wen, S., Zhao, Y., Zhang, H., Lam, H., & Manfredi, L. (2020) · Autonomous Robots · doi:10.1007/s10514-019-09897-6 · Cited by 14
“Newcombe et al. proposed the dense tracking and mapping method called DTAM [7], Forster et al. proposed fast semi-direct monocular visual odometry methods called SVO [8], Engle et al. proposed LSD-SLAM [9] and DSO [4].”
MethodSLAM-Based Multistate Tracking System for Mobile Human-Robot Interaction
Hempel, T., & Al-Hamadi, A. K. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-50347-5_32 · Cited by 14
Zhao, Y., Liu, G., Xu, S., Bu, S., Jiang, H., & Wan, G. (2020) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2020.3008517 · Cited by 16
“SLAM technology has rapidly developed in recent years, and monocular visual SLAM systems, including keypoint-based [34]–[36], direct [37]–[39], and semidirect methods [40], [41], have been proposed.”
BackgroundDistributed Photometric Bundle Adjustment
Demmel, N., Gao, M., Laude, E., Wu, T. T., & Cremers, D. (2020) · International Conference on 3D Vision (3DV) · doi:10.1109/3dv50981.2020.00024 · Cited by 14
“In the last decade, direct methods relying on the photometric error have become popular for pose estimation (typically frame-to-frame) [18, 9] and (semi-)dense depth estimation [9, 20, 25].”
MethodDetecting Markers in Blurred and Defocused Images
Toyoura, M., Aruga, H., Turk, M., & Mao, X. (2013) · International Conference on Cyberworlds (CW) · doi:10.1109/cw.2013.58 · Cited by 14
“The feature points are detected not in planar markers but in general objects.”
BackgroundFeature-based or Direct: An Evaluation of Monocular Visual Odometry.Preprint
Yang, N., Wang, R., & Cremers, D. (2017) · arXiv (Cornell University) · Cited by 14
DetectFusion: Detecting and Segmenting Both Known and Unknown Dynamic Objects in Real-time SLAMPreprint
Hachiuma, R., Pirchheim, C., Schmalstieg, D., & Saitô, H. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1907.09127 · Cited by 17
“Therefore, we make the following contributions: Instance segmentation at full frame rate By detecting and segmenting known object instances in each incoming RGB-D frame in real-time (we reach about 20 fps including tracking and mapping), we can create new object maps faster (just-in-time), update multiple maps more ac…”
InfluentialTrained Trajectory based Automated Parking System using Visual SLAM.Preprint
Tripathi, N., Sistu, G., & Yogamani, S. K. (2020) · arXiv (Cornell University) · Cited by 15
Real-time image-based RGB-D camera motion tracking and environment mapping
Tykkälä, T. (2013) · theses.fr (ABES) · doi:10.70675/f735edd2ze710z4718z992bz4ffb9a5b780f · Cited by 17
Easi3R: Estimating Disentangled Motion from DUSt3R Without Training
Chen, X., Chen, Y., Xiu, Y., Geiger, A., & Chen, A. (2025) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51701.2025.00856 · Cited by 102
“Structure-from-Motion (SfM) [2, 40, 41, 47, 50, 51] and Simultaneous Localization and Mapping (SLAM) [9, 13, 32, 34] have long been the foundation for 3D structure and camera pose estimation.”
A review on monocular tracking and mapping: from model-based to data-driven methods
Gadipudi, N., Elamvazuthi, I., Izhar, L. I., Tiwari, L., Hebbalaguppe, R., Lu, C., & Doss, A. S. A. (2022) · The Visual Computer · doi:10.1007/s00371-022-02702-z · Cited by 17
“Considering the drawbacks of the high computational resources used in [98], and sparse map generation in [70], Engel et al. [99] proposed a direct technique by using only the pixels with high gradients while ignoring the uniform areas with PGO.”
MethodConstructing Maps for Autonomous Robotics: An Introductory Conceptual Overview
Racinskis, P., Ārents, J., & Greitāns, M. (2023) · Electronics · doi:10.3390/electronics12132925 · Cited by 19
DZ-SLAM: A SAM-based SLAM algorithm oriented to dynamic environments
Chen, Z., Zang, Q., & Zhang, K. (2024) · Displays · doi:10.1016/j.displa.2024.102846 · Cited by 16
Dense mono reconstruction: Living with the pain of the plain plane
Piniés, P., Paz, L. M., & Newman, P. M. (2015) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra.2015.7139927 · Cited by 13
“We have tested our own CUDA implementation of the previous (TV) algorithm on a NVIDIA GeForce GT 650M 1024 MB card using different synthetic scenarios found in [8, 9] obtaining a median depthmap error that is usually below 2 centimetres, after 800 iterations in 500 ms.”
BackgroundMethodBecker, F., Lenzen, F., Kappes, J. H., & Schnörr, C. (2013) · International Journal of Computer Vision · doi:10.1007/s11263-013-0639-7 · Cited by 14
“Dense methods for egomotion estimation from monocular (Newcombe et al. 2011; Sheikh et al. 2007) or stereo data (Comport et al.”
BackgroundMethodHuman-Like Sensing for Robotic Remote Inspection and Analytics
Pal, A., Dasgupta, R., Saha, A., & Nandi, B. (2016) · Wireless Personal Communications · doi:10.1007/s11277-016-3239-3 · Cited by 13
“The 3D reconstruction of an environment from multiple images or video captured by a single moving camera has been studied for several years and is well known as Structure-from-Motion (SfM) and quite powerful SfM pipeline readily available for 3D reconstructing as shown in [9, 10].”
MethodFrom Local Understanding to Global Regression in Monocular Visual Odometry
Esfahani, M. A., Wu, K., Yuan, S., & Wang, H. (2019) · International Journal of Pattern Recognition and Artificial Intelligence · doi:10.1142/s0218001420550022 · Cited by 16
A Low-cost Simultaneous Localization And Mapping Algorithm For Last-mile Indoor Delivery
Wang, W., Zhao, W., Wang, X., Jin, Z., Li, Y., & Runge, T. M. (2019) · International Conference on Transportation Information and Safety · doi:10.1109/ictis.2019.8883749 · Cited by 23
Naudet-Collette, S., Melbouci, K., Gay‐Bellile, V., Ait-Aider, O., & Dhome, M. (2020) · Robotica · doi:10.1017/s0263574720000363 · Cited by 13
DOE-SLAM: Dynamic Object Enhanced Visual SLAM
Hu, X., & Lang, J. (2021) · Sensors · doi:10.3390/s21093091 · Cited by 11
“Direct SLAM (e.g., LSD-SLAM [17] and DTAM [18]) uses all the information from the sensor without pre-processing.”
MethodStereo Visual Odometry Pose Correction through Unsupervised Deep Learning
Zhang, S., Lu, S., He, R., & Bao, Z. (2021) · Sensors · doi:10.3390/s21144735 · Cited by 14
“Significantly, several state-of-the-art VO systems have been designed based on feature point matching [1–3] and constant gray hypothesis [4–6].”
BackgroundContextAVO: Local context guided and refining poses for deep visual odometry
Song, R., Zhu, R., Xiao, Z., & Yan, B. (2023) · Neurocomputing · doi:10.1016/j.neucom.2023.02.014 · Cited by 14
Too much TV is bad: Dense reconstruction from sparse laser with non-convex regularisation
Piniés, P., Paz, L. M., & Newman, P. M. (2015) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra.2015.7138991 · Cited by 13
“A consequence of our analysis and optimisation construction is the conclusion that new more complex norms should be given a central role in dense reconstruction of large outdoor scenes - in contrast to the usual standard L1 and Huber norms successfully applied in small environments [2], [3].”
ResultGao, Q., Chang, P.-L., Rueckert, D., Ali, S. M., Cohen, D., Pratt, P., Mayer, E. K., Yang, G., Darzi, A. W., & Edwards, P. (2013) · Computerized Medical Imaging and Graphics · doi:10.1016/j.compmedimag.2013.01.001 · Cited by 13
Visual–Inertial Combined Odometry System for Aerial Vehicles
Zhang, J., & Singh, S. (2015) · Journal of Field Robotics · doi:10.1002/rob.21599 · Cited by 15
Robust visual odometry to irregular illumination changes with RGB-D camera
Kim, P., Lim, H., & Kim, H. J. (2015) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros.2015.7353893 · Cited by 15
“However, direct methods are getting more interests recently [3], [9], [10].”
Air-SSLAM: A Visual Stereo Indoor SLAM for Aerial Quadrotors
Araujo, P., Miranda, R., do Carmo, D. S., Alves, R., & de Oliveira, L. R. (2017) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2017.2730883 · Cited by 13
Kaipa, K. N., Morato, C. W., & Gupta, S. K. (2018) · Journal of Computing and Information Science in Engineering · doi:10.1115/1.4039061 · Cited by 13
Real-time dense map fusion for stereo SLAM
Pire, T., Baravalle, R., D'Alessandro, A., & Civera, J. (2018) · Robotica · doi:10.1017/s0263574718000528 · Cited by 15
“g.(22,27) Our algorithm, based on fusing dense stereo depth maps, is able to estimate dense local maps from stereo in real time on a standard CPU.”
MethodEvaluation Criteria for Inside-Out Indoor Positioning Systems Based on Machine Learning
Löffler, C., Riechel, S., Fischer, J., & Mutschler, C. (2018) · International Conference on Indoor Positioning and Indoor Navigation · doi:10.1109/ipin.2018.8533862 · Cited by 18
“The construction of scene models through dense and sub-pixel multi-view stereo reconstruction [12], [13] using direct image alignment, or generative models [14], i.”
MethodSparse Representations for Object- and Ego-Motion Estimations in Dynamic Scenes
Kashyap, H. J., Fowlkes, C. C., & Krichmar, J. L. (2020) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2020.3006467 · Cited by 14
Real-Time 3D Reconstruction Method Based on Monocular Vision
Jia, Q., Chang, L., Qiang, B., Zhang, S., Xie, W., Yang, X., Sun, Y., & Yang, M. (2021) · Sensors · doi:10.3390/s21175909 · Cited by 16
“proposed a real‐time 3D reconstruction system based on multi‐view constraints [17].”
BackgroundSingh, A., Narula, R., Rashwan, H. A., Abdel‐Nasser, M., Puig, D., & Nandi, G. C. (2022) · Neural Computing and Applications · doi:10.1007/s00521-022-07273-7 · Cited by 16
“A direct technique is employed in the technology known as DTAM (Dense Tracking and Mapping In Real Life) [33].”
MethodContext-Aware Mixed Reality: A Framework for Ubiquitous InteractionPreprint
Chen, L. F., Tang, W. J., John, N. W., Wan, T. R., & Zhang, J. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1803.05541 · Cited by 13
“Our approach draws on recent success of dense SLAM algorithms [4] [5] [6] and deep learning for semantic seg-mentations [8] [9] [10] [11] [12] that have been mostly used in the field of robotics until now.”
BackgroundMethodInfluentialActive 3D Motion Visualization Based on Spatiotemporal Light-Ray Integration
Sakaue, F., & Sato, J. (2020) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr42600.2020.00205 · Cited by 12
“In computer vision, 3D motion estimation has a long history, and many efficient methods have been proposed under various conditions [17, 4, 13, 14, 6].”
MethodSystematic literature review of collaborative SLAM applied to autonomous mobile robots
Aloui, K., Guizani, A., Hammadi, M., Haddar, M., & Soriano, T. (2022) · IEEE Information Technologies & Smart Industrial Systems (ITSIS) · doi:10.1109/itsis56166.2022.10118378 · Cited by 13
Efficient Map Fusion for Multiple Implicit SLAM Agents
Liu, S., & Zhu, J. (2023) · IEEE Transactions on Intelligent Vehicles · doi:10.1109/tiv.2023.3297194 · Cited by 19
“Traditional SLAM methods use feature extraction, matching [1], [2], [5], or nonlinear energy minimization [3], [7] across different video frames in the frontend and convert those feature tracks into camera poses and 3D coordinates based on a filter or a nonlinear least squares optimizer in the…”
BackgroundMethodDeepSFM: Robust Deep Iterative Refinement for Structure From Motion
Ren, X., Wei, X., Li, Z., Fu, Y., Zhang, Y., & Xue, X. (2023) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2023.3307567 · Cited by 19
“Simultaneously estimating 3 d structure and camera motion is a well studied problem which has a traditional tool-chain of techniques [11], [38], [39].”
BackgroundMOLO-SLAM: A Semantic SLAM for Accurate Removal of Dynamic Objects in Agricultural Environments
Lv, J., Yao, B., Guo, H., Gao, C., Wu, W., Li, J., Sun, S., & Luo, Q. (2024) · Agriculture · doi:10.3390/agriculture14060819 · Cited by 16
Wenzel, P., Yang, N., Wang, R., Zeller, N., & Cremers, D. (2024) · International Journal of Computer Vision · doi:10.1007/s11263-024-02230-4 · Cited by 24
“During the last decade, research on visual odometry (VO) and simultaneous localization and mapping (SLAM) has made tremendous strides [13, 16, 40, 41], particularly in the context of autonomous driving [14, 39, 69, 73].”
Robust dense visual odometry for RGB-D cameras in a dynamic environment
Dib, A., & Charpillet, F. (2015) · International Conference on Advanced Robotics (ICAR) · doi:10.1109/icar.2015.7298210 · Cited by 16
“It has been shown in [24] and [25] that dense methods outperform feature-based.”
Visual Odometry and Place Recognition Fusion for Vehicle Position Tracking in Urban Environments
Ouerghi, S., Boutteau, R., Savatier, X., & Tlili, F. (2018) · Sensors · doi:10.3390/s18040939 · Cited by 12
“Simultaneous Localization and Mapping (SLAM) methods attempt to reduce this drift by using landmarks and jointly optimizing over all or a selection of poses and landmarks [18,19].”
MethodHybrid-Residual-Based RGBD Visual Odometry
Yu, Q., Xiao, J., Lu, H., & Zheng, Z. (2018) · IEEE Access · doi:10.1109/access.2018.2836928 · Cited by 13
“According to the number of pixels, the methods can be divided into dense methods [7] and sparse methods [8], [9].”
MethodTowards Dynamic Monocular Visual Odometry Based on an Event Camera and IMU Sensor
Mohamed, S. A. S., Haghbayan, M.-H., Rabah, M., Heikkonen, J., Tenhunen, H., & Plosila, J. (2020) · Lecture notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering · doi:10.1007/978-3-030-38822-5_17 · Cited by 12
“Direct approaches analyze the intensity of pixels in the image to recover the pose of the camera [12– 14].”
MethodLegged Robots for Autonomous Inspection and Monitoring of Offshore Assets
Ramezani, M., Brandão, M., Casseau, B., Havoutis, I., & Fallon, M. F. (2020) · Offshore Technology Conference · doi:10.4043/30694-ms · Cited by 12
Monocular Visual Odometry using Learned Repeatability and Description
Huang, H., Ye, H., Sun, Y., & Liu, M. (2020) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra40945.2020.9197406 · Cited by 13
“Direct methods [21](cid:150)[23], on the contrary, directly minimize the photometric error on input images to track camera poses.”
MethodPhotometric single-view dense 3D reconstruction in endoscopy
Batlle, V. M., Montiel, J. M. M., & Tardós, J. D. (2022) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros47612.2022.9981742 · Cited by 17
“Other authors focus on the study of photometry to obtain dense and semi-dense reconstructions of outdoor and indoor 3D scenes [10], [11].”
BackgroundMethodPVI-DSO: Leveraging Planar Regularities for Direct Sparse Visual-Inertial Odometry
Xu, B., Li, X., Wang, J., Yuen, C., & Li, J. (2023) · IEEE Sensors Journal · doi:10.1109/jsen.2023.3285786 · Cited by 15
“The dense mapping methods [15], [16] based on monocular vision and pixel-wise reconstruction lead to high time complexity and loss accuracy due to decoupling the trajectory estimation and mapping.”
MethodÁlvarez-Tuñón, O., Brodskiy, Y., & Kayacan, E. (2023) · IEEE Transactions on Artificial Intelligence · doi:10.1109/tai.2023.3321032 · Cited by 23
“2) Direct Methods: Direct methods, introduced in 2007 by DTAM [33], use image intensities rather than sparse features.”
BackgroundMethodInfluentialDense 3D Reconstruction with a Hand-Held Camera
Ummenhofer, B., & Brox, T. (2012) · Lecture notes in computer science · doi:10.1007/978-3-642-32717-9_11 · Cited by 12
“Newcombe et al. [14] presented an interactive approach that allows for acquisition of dense 3D models in real-time.”
BackgroundMethodVisual recovery of saliency maps from human attention in 3D environments
Santner, K., Fritz, G., Paletta, L., & Mayer, H. (2013) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra.2013.6631185 · Cited by 12
“They extended their approach without relying on sparse features and key frames named DTAM [8], where photometric information is fused into a single cost volume to estimate the depth map for selected key frames.”
BackgroundBayesian filtering for keyframe-based visual SLAM
Kim, J., Yoon, K., & Kweon, I. S. (2014) · The International Journal of Robotics Research · doi:10.1177/0278364914550215 · Cited by 12
“Klein and Murray (2007) presented the PTAM algorithm to produce detailed maps with many landmarks by separating tracking and mapping operations, and Newcombe et al. (2011) proposed a system for real-time camera tracking and dense reconstruction using a single camera.”
BackgroundReal-time sequential model-based non-rigid SFM
Bronte, S., Paladini, M., Bergasa, L. M., Agapito, L., & Arroyo, R. (2014) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros.2014.6942684 · Cited by 12
“Simultaneous reconstruction and tracking for rigid objects is a mature area in which real-time methods have been developed using sparse features [5] and dense maps [6], [7].”
BackgroundMethodMonocular Visual Odometry and Dense 3D Reconstruction for On-Road Vehicles
Zhu, M., Ramalingam, S., Taguchi, Y., & Garaas, T. W. (2012) · Lecture notes in computer science · doi:10.1007/978-3-642-33868-7_59 · Cited by 12
“Several SLAM techniques for general motion sequences exist in the robotics community [9,13].”
BackgroundLarge-Scale Dense 3D Reconstruction from Stereo Imagery
Alcantarilla, P. F., Beall, C., & Dellaert, F. (2013) · Cited by 12
Detecting, localizing, and recognizing trees with a monocular MAV: Towards preventing deforestation
Shah, U., Khawad, R., & Krishna, K. M. (2017) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra.2017.7989229 · Cited by 12
“Moreover, DTAM is susceptible to breakage due to insufficient feature tracks if MAV pitches or rolls significantly.”
MethodLAP-SLAM: A Line-Assisted Point-Based Monocular VSLAM
Zhang, F., Rui, T., Yang, C., & Shi, J. (2019) · Electronics · doi:10.3390/electronics8020243 · Cited by 12
“These techniques can produce a dense [1] or semi-dense [2] reconstruction of the scene.”
BackgroundMethodFlow-based SLAM: From geometry computation to learning
Yan, Z., & Zha, H. (2019) · Virtual Reality & Intelligent Hardware · doi:10.1016/j.vrih.2019.09.001 · Cited by 12
Edge-based cover recognition and tracking method for an AR-aided aircraft inspection system
Yang, X., Fan, X., Wang, J., Yin, X., & Qiu, S. (2020) · The International Journal of Advanced Manufacturing Technology · doi:10.1007/s00170-020-06301-x · Cited by 12
“DTAM [38] is a novel approach to dense visual odometry through extensive GPGPU parallelization.”
MethodLighting Enhancement Aids Reconstruction of Colonoscopic Surfaces
Zhang, Y., Wang, S., Ma, R., McGill, S., Rosenman, J. G., & Pizer, S. M. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-78191-0_43 · Cited by 13
UNVEILING LARGE-SCALE HISTORICAL CONTENTS WITH V-SLAM AND MARKERLESS MOBILE AR SOLUTIONS
Torresani, A., Rigon, S., Farella, E. M., Menna, F., & Remondino, F. (2021) · The international archives of the photogrammetry, remote sensing and spatial information sciences/International archives of the photogrammetry, remote sensing and spatial information sciences · doi:10.5194/isprs-archives-xlvi-m-1-2021-761-2021 · Cited by 16
“DTAM (Newcombe et al., 2011) and LSDSLAM (Engel et al., 2014) are well known direct algorithms, 4 https://totem.fbk.eu/ 5 https://judit.fbk.eu/ This contribution has been peer-reviewed. https://doi.org/10.5194/isprs-archives-XLVI-M-1-2021-761-2021 | © Author(s) 2021.”
BackgroundInfluentialPan, Z., Hou, J., & Yu, L. (2022) · Measurement Science and Technology · doi:10.1088/1361-6501/ac505b · Cited by 14
“In this paper, optimized pose information is used to refuse the RGB-D data stream [29, 30].”
MethodExperimentation and Simulation with Autonomous Coverage Path Planning for UAVs
Biundini, I. Z., Melo, A. G., Coelho, F. O., Honório, L. D. M., Marcato, A. L. M., & Pinto, M. F. (2022) · Journal of Intelligent & Robotic Systems · doi:10.1007/s10846-022-01654-9 · Cited by 12
“[42, 43] present optimization proposals separate from the photometric errors and the mapping issue to generate the best quality of the two objectives.”
Background
Metadata
- Authors
- Newcombe, R. A., Lovegrove, S. J., & Davison, A. J.
- Affiliations
- Imperial College London (GB)
- ORCID
- Davison, Andrew J.
- Venue
- 2011 International Conference on Computer Vision
- Publisher
- IEEE
- Published
- 2011-11
- Pages
- 2320-2327
- OpenAlex
- W2108134361
- Cited by
- 2,078
- References
- 17
- Citations by year
- 2026: 49 · 2025: 144 · 2024: 163 · 2023: 140 · 2022: 141 · 2021: 166 · 2020: 179 · 2019: 199 · 2018: 183 · 2017: 176 · 2016: 136 · 2015: 151 · 2014: 102 · 2013: 92 · 2012: 51
- FWCI
- 154.56
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- IEEE/CVF International Conference on Computer Vision (ICCV) (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Robotics and Sensor-Based Localization
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Vision and Imaging
- Physical Sciences › Earth and Planetary Sciences › Geology › 3D Surveying and Cultural Heritage
- Keywords
- dense 3D reconstruction, real-time mapping, depth images, non-convex optimization, energy minimization, image registration, GPU parallelization, augmented reality, handheld camera
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.