Multimodal scale estimation for monocular visual odometry

Fanani, Nolang; Sturck, Alina; Barnada, Marc; Mester, Rudolf

doi:10.1109/ivs.2017.7995955

Cited by 17 publications

(16 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In order to provide meaningful information to the network, the input cues are scale-aware. This can be easily obtained at training time by running a stereo VO algorithm [31], while at test time a monocular VO framework with any scale recovery, as for instance [43,12], is required.…”

Section: Methodsmentioning

confidence: 99%

“…Large progress has been achieved in the development of VO and SLAM methods [9,10,31,32]. Although a stereo setup [9,11,31] avoids scale ambiguity, recent trend aims at recovering the scale of monocular VO exploiting geometry [43,12] or deep learning [39,47,46].…”

Section: Related Workmentioning

confidence: 99%

“…The two respectively exploit stereo and monocular sequences. While the former provides the correct scale, the second requires a scale recovery mechanisms as for instance in [43,12]. In order to ease the learning process, we perform a first round of training on K r + K o feeding the network with the 3D points from stereo VO.…”

Section: Implementation Detailsmentioning

confidence: 99%

See 2 more Smart Citations

Enhancing Self-Supervised Monocular Depth Estimation with Traditional Visual Odometry

Andraghetti¹,

Myriokefalitakis²,

Dovesi³

et al. 2019

2019 International Conference on 3D Vision (3DV)

View full text Add to dashboard Cite

Estimating depth from a single image represents an attractive alternative to more traditional approaches leveraging multiple cameras. In this field, deep learning yielded outstanding results at the cost of needing large amounts of data labeled with precise depth measurements for training. An issue softened by self-supervised approaches leveraging monocular sequences or stereo pairs in place of expensive ground truth depth annotations. This paper enables to further improve monocular depth estimation by integrating into existing self-supervised networks a geometrical prior. Specifically, we propose a sparsity-invariant autoencoder able to process the output of conventional visual odometry algorithms working in synergy with depth-from-mono networks. Experimental results on the KITTI dataset show that by exploiting the geometrical prior, our proposal: i) outperforms existing approaches in the literature and ii) couples well with both compact and complex depth-from-mono architectures, allowing for its deployment on high-end GPUs as well as on embedded devices (e.g., NVIDIA Jetson TX2).

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Enhancing Self-Supervised Monocular Depth Estimation with Traditional Visual Odometry

Andraghetti¹,

Myriokefalitakis²,

Dovesi³

et al. 2019

2019 International Conference on 3D Vision (3DV)

View full text Add to dashboard Cite

show abstract

“…In this paper, a good ground point should be in the center bottom area of the image, present a good image gradient along the epipolar line to enhance matching precision and is expected to be close to the estimated ground. A multimodal mechanism of prediction, classification, and correction have been proposed in [36]. For robust estimation, the scale correction scheme combines cues from both dense and sparse ground plane estimation.…”

Section: B Monocular-based Methodsmentioning

confidence: 99%

Ground-Plane-Based Absolute Scale Estimation for Monocular Visual Odometry

Zhou

Dai

2020

IEEE Trans. Intell. Transport. Syst.

View full text Add to dashboard Cite

Recovering absolute metric scale from a monocular camera is a challenging but highly desirable problem for monocular camera-based systems. By using different kinds of cues, various approaches have been proposed for scale estimation, such as camera height, object size etc. In this paper, firstly, we summarize different kinds of scale estimation approaches. Then, we propose a robust divide and conquer absolute scale estimation method based on the ground plane and camera height by analyzing the advantages and disadvantages of different approaches. By using the estimated scale, an effective scale correction strategy has been proposed to reduce the scale drift during the Monocular Visual Odometry (VO) estimation process. Finally, the effectiveness and robustness of the proposed method have been verified on both public and self-collected image sequences.

show abstract

“…For example, a fixed height of the camera above the ground plane is useful to estimate scale and therefore avoid scale drift [10], [11], [12], [13]. Alternatively, the size of known objects in the environment can be used as a depth cue [14], [15].…”

Section: A Scale Drift On Monocular Slammentioning

confidence: 99%

Eliminating Scale Drift in Monocular SLAM Using Depth From Defocus

Shiozaki

Dissanayake

2018

IEEE Robot. Autom. Lett.

View full text Add to dashboard Cite

Abstract-This paper presents a novel approach to correct errors caused by accumulated scale drift in monocular SLAM. It is shown that the metric scale can be estimated using information gathered through monocular SLAM and image blur due to defocus. A nonlinear least squares optimization problem is formulated to integrate depth estimates from defocus to monocular SLAM. An algorithm to process the output keyframe and feature location estimates generated by a monocular SLAM algorithm to correct for scale drift at selected local regions of the environment is presented. The proposed algorithm is experimentally evaluated by processing the output of ORB-SLAM [1] to obtain accurate metric scale maps from a monocular camera without any prior knowledge about the scene.

show abstract

Multimodal scale estimation for monocular visual odometry

Cited by 17 publications

References 18 publications

Enhancing Self-Supervised Monocular Depth Estimation with Traditional Visual Odometry

Enhancing Self-Supervised Monocular Depth Estimation with Traditional Visual Odometry

Ground-Plane-Based Absolute Scale Estimation for Monocular Visual Odometry

Eliminating Scale Drift in Monocular SLAM Using Depth From Defocus

Contact Info

Product

Resources

About