Monocular depth estimation for 2D mapping of simulated environments
https://doi.org/10.17586/2226-1494-2024-24-1-118-123
Abstract
This article addresses the problem of constructing maps for 2D simulated environments. An algorithm based on monocular depth estimation is proposed achieving comparable accuracy to methods utilizing expensive sensors such as RGBD cameras and LIDARs. To solve the problem, we employ a multi-stage approach. First, a neural network predicts a relative disparity map from an RGB flow provided by RGBD camera. Using depth measurements from the same camera, two parameters are estimated that connect the relative and absolute displacement maps in the form of a linear regression relation. Based on a simpler RGB camera, by applying a neural network and estimates of scaling parameters, an estimate of the absolute displacement map is formed, which allows to obtain an estimate of the depth map. Thus, a virtual scanner has been designed providing Cartographer SLAM with depth information for environment mapping. The proposed algorithm was evaluated on a ROS 2.0 simulation of a simple mobile robot. It achieves faster depth prediction compared to other depth estimation algorithms. Furthermore, maps generated by our approach demonstrated a high overlap ratio with those obtained using an ideal RGBD camera. The proposed algorithm can find applicability in crucial tasks for mobile robots, like obstacle avoidance, and path planning. Moreover, it can be used to generate accurate cost maps, enhancing safety and adaptability in mobile robot navigation.
Supporting agencies: This paper was supported by the Ministry of Science and Higher Education of the Russian Federation (State Assignement No. 2019-0898).
About the Authors
M. Barhoum
ITMO University
Russian Federation
Russian Federation
Majd Barhoum — PhD Student
Saint Petersburg, 197101
A. A. Pyrkin
ITMO University
Russian Federation
Russian Federation
Anton A. Pyrkin — D.Sc., Full Professor
Saint Petersburg, 197101
sc 26656070700
References
1. Bhat S.F., Alhashim I., Wonka P. AdaBins: Depth estimation using adaptive bins. Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 4009–4018. https://doi.org/10.1109/cvpr46437.2021.00400
2. Li Z., Wang X., Liu X., Jiang J. BinsFormer: Revisiting adaptive bins for monocular depth estimation. arXiv, 2022, arXiv:2204.00987. https://doi.org/10.48550/arXiv.2204.00987
3. Zhang S., Yang L., Mi M.B., Zheng X., Yao A. Improving deep regression with ordinal entropy. arXiv, 2023, arXiv:2301.08915. https://doi.org/10.48550/arXiv.2301.08915
4. Ranftl R., Bochkovskiy A., Koltun V. Vision transformers for dense prediction. Proc. of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 12179–12188. https://doi.org/10.1109/iccv48922.2021.0119
5. Xie Z., Geng Z., Hu J., Zhang Z., Hu H., Cao Y. Revealing the dark secrets of masked image modeling. Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 14475–14485. https://doi.org/10.1109/cvpr52729.2023.01391
6. Birkl R., Wofk D., Müller M. MiDaS v3.1 – A Model zoo for robust monocular relative depth estimation. arXiv, 2023, arXiv:2307.14460. https://doi.org/10.48550/arXiv.2307.14460
7. Bhat S.F., Birkl R., Wofk D., Wonka P., Müller M. ZoeDepth: Zero-shot transfer by combining relative and metric depth. arXiv, 2023, arXiv:2302.12288. https://doi.org/10.48550/arXiv.2302.12288
8. Ranftl R., Lasinger K., Hafner D., Schindler K., Koltun V. Towards robust monocular depth estimation: mixing datasets for zero-shot cross-dataset transfer. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2022, vol. 44, no. 3, pp. 1623–1637. https://doi.org/10.1109/tpami.2020.3019967
9. Xing X., Cai Y., Lu T., Yang Y., Wen D. Joint self-supervised monocular depth estimation and SLAM. Proc. of the 26th International Conference on Pattern Recognition (ICPR), 2022, pp. 4030–4036. https://doi.org/10.1109/icpr56361.2022.9956576
10. Geng M., Shang S., Ding B., Wang H., Zhang P. Unsupervised learning-based depth estimation-aided visual slam approach. Circuits, Systems, and Signal Processing, 2020, vol. 39, pp. 543–570. https://doi.org/10.1007/s00034-019-01173-3
11. Li Z., Snavely N. MegaDepth: Learning single-view depth prediction from internet photos. Proc. of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2018, pp. 2041–2050. https://doi.org/10.1109/cvpr.2018.00218
12. Tran M., Ly N. Mobile robot planner with low-cost cameras using deep reinforcement learning. Proc. of the 7th NAFOSTED Conference on Information and Computer Science (NICS), 2020, pp. 54–59. https://doi.org/10.1109/nics51282.2020.9335852
13. Hess W., Kohler D., Rapp H., Andor D. Real-time loop closure in 2D LIDAR SLAM. Proc. of the 2016 IEEE International Conference on Robotics and Automation (ICRA), 2016, pp. 1271–1278. https://doi.org/10.1109/icra.2016.7487258
14. Eigen D., Puhrsch C., Fergus R. Depth map prediction from a single image using a multi-scale deep network. Advances in Neural Information Processing Systems, 2014, vol. 27, pp. 2, 5, 6.
15. Garg R., Kumar B.G.V., Carneiro G., Reid I. Unsupervised CNN for single view depth estimation: Geometry to the rescue. Lecture Notes in Computer Science, 2016, vol. 9912, pp. 740–756. https://doi.org/10.1007/978-3-319-46484-8_45
Review
For citations:
Barhoum M., Pyrkin A.A. Monocular depth estimation for 2D mapping of simulated environments. Scientific and Technical Journal of Information Technologies, Mechanics and Optics. 2024;24(1):118-123. https://doi.org/10.17586/2226-1494-2024-24-1-118-123
Views:
10
Email this article 