Comparative Analysis of Monocular SLAM Algorithms Using TUM and EuRoC Benchmarks

Mingachev, Eldar; Lavrenov, Roman; Magid, Evgeni

doi:10.1007/978-981-15-5580-0_28

Cited by 5 publications

(5 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the previous studies of [39,40], the authors compared the DSO, LDSO, ORB-SLAM2, and DynaSLAM algorithms on the same TUM-Mono benchmark, and their findings mostly matched what we observed during this evaluation. However, we extended their study considerably by implementing six additional methods following the taxonomy described in [9] and performed an appropriate statistical analysis to determine the significant differences in each system's performance.…”

Section: Discussionsupporting

confidence: 76%

“…The DSM performed similarly to the of the sparse-direct approaches but occasionally presented trajectory loss issues affec the overall performance. Furthermore, the DynaSLAM considerably outperformed ORB-SLAM2, especially in challenging sequences like 18,19,21,22,23,27,28,38,39,40, among others, where the ORB-SLAM commonly failed. However, it occasionally sented trajectory loss and initialization issues.…”

Section: Comparative Analysismentioning

confidence: 99%

“…In contrast, we compared ten methods following a taxonomy, where the three methods tested in [38] were addressed, exploring the same photometric calibration, motion bias, and rolling shutter effects by applying the TUM-Mono dataset. Then in 2020, Mingachev et al published two comparisons [39,40] testing first the DSO, LDSO, and ORB-SLAM2 and then the ROS-based methods, DSO, LDSO, DynaSLAM, and ORB-SLAM2, on the TUM-Mono and EuRoC benchmarks, where the authors verified the performance of the algorithms implementing an open-source code in their hardware to determine whether there were improvements in the LDSO and DynaSLAM-updates of the original DSO and ORB-SLAM2. The authors found that the updates achieved slight error reductions over their predecessors on both benchmarks, reported as medians of 10 executions of each algorithm in each sequence.…”

Section: Comparisonsmentioning

confidence: 99%

“…As can be noticed, the most complete and largest dataset of the above is the TUM-Mono, which is why this dataset was applied in this comparison study. It also has the advantage of being the only dataset that was obtained purely depending on a monocular RGB setup, without depending on any additional sensor or source of information as mentioned in [6,39,40], making it ideal for comparing visual-only SLAM and VO systems. In addition, this benchmark provides the most complete set of metrics that can be explored to efficiently compare the selected algorithms in multiple dimensions-discussed in the following section.…”

mentioning

confidence: 99%

See 3 more Smart Citations

A Comparison of Monocular Visual SLAM and Visual Odometry Methods Applied to 3D Reconstruction

Herrera-Granda,

Torres-Cantero,

Rosales

et al. 2023

Applied Sciences

View full text Add to dashboard Cite

Pure monocular 3D reconstruction is a complex problem that has attracted the research community’s interest due to the affordability and availability of RGB sensors. SLAM, VO, and SFM are disciplines formulated to solve the 3D reconstruction problem and estimate the camera’s ego-motion; so, many methods have been proposed. However, most of these methods have not been evaluated on large datasets and under various motion patterns, have not been tested under the same metrics, and most of them have not been evaluated following a taxonomy, making their comparison and selection difficult. In this research, we performed a comparison of ten publicly available SLAM and VO methods following a taxonomy, including one method for each category of the primary taxonomy, three machine-learning-based methods, and two updates of the best methods to identify the advantages and limitations of each category of the taxonomy and test whether the addition of machine learning or updates on those methods improved them significantly. Thus, we evaluated each algorithm using the TUM-Mono dataset and benchmark, and we performed an inferential statistical analysis to identify the significant differences through its metrics. The results determined that the sparse-direct methods significantly outperformed the rest of the taxonomy, and fusing them with machine learning techniques significantly enhanced the geometric-based methods’ performance from different perspectives.

show abstract

Section: Discussionsupporting

confidence: 76%

Section: Comparative Analysismentioning

confidence: 99%

Section: Comparisonsmentioning

confidence: 99%

mentioning

confidence: 99%

See 2 more Smart Citations

A Comparison of Monocular Visual SLAM and Visual Odometry Methods Applied to 3D Reconstruction

Herrera-Granda,

Torres-Cantero,

Rosales

et al. 2023

Applied Sciences

View full text Add to dashboard Cite

show abstract

Section: Discussionsupporting

confidence: 76%

Comparison of Monocular Visual SLAM and Visual Odometry Methods Applied to 3D Reconstruction

Herrera-Granda¹,

Torres²,

Rosales³

et al. 2023

Preprint

View full text Add to dashboard Cite

Pure monocular 3D reconstruction is an ill-posed problem that has attracted the research community's interest due to the affordability and availability of RGB sensors. SLAM, VO, and SFM are disciplines formulated to solve the 3D reconstruction problem and estimate the camera’s ego-motion, so many methods have been proposed. However, most of these methods were not evaluated in large datasets, under various motion patterns, had not been tested under the same metrics, and most of them had not been evaluated following a taxonomy, making their comparison and selection difficult. In this research, we performed a comparison of ten publicly available SLAM and VO methods following a taxonomy, including one method for each category of the primary taxonomy, three machine learning-based methods, and two updates of the best methods, to identify the advantages and limitations of each category of the taxonomy and test if the addition of machine learning or the updates made on those methods improved them significantly. Thus, we evaluated each algorithm under the TUM-Mono benchmark and performed an inferential statistical analysis to identify significative differences through its metrics. Results determined that sparse-direct methods significantly outperformed the rest of the taxonomy, and fusing them with machine learning techniques significantly improves the performance of geometric-based methods from different perspectives.

show abstract

Comparison of Monocular ROS-Based Visual SLAM Methods

Apurin

Abbyasov

Tsoy

et al. 2022

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Comparative Analysis of Monocular SLAM Algorithms Using TUM and EuRoC Benchmarks

Cited by 5 publications

References 18 publications

A Comparison of Monocular Visual SLAM and Visual Odometry Methods Applied to 3D Reconstruction

A Comparison of Monocular Visual SLAM and Visual Odometry Methods Applied to 3D Reconstruction

Comparison of Monocular Visual SLAM and Visual Odometry Methods Applied to 3D Reconstruction

Comparison of Monocular ROS-Based Visual SLAM Methods

Contact Info

Product

Resources

About