novtexmechМехатроника, автоматизация, управлениеMekhatronika, Avtomatizatsiya, Upravlenie1684-64272619-1253Commercial Publisher «New Technologies»10.17587/mau.27.59-65novtexmech-1920Research ArticleРОБОТЫ, МЕХАТРОНИКА И РОБОТОТЕХНИЧЕСКИЕ СИСТЕМЫROBOT, MECHATRONICS AND ROBOTIC SYSTEMSАнализ робастности метода неявного представления карт в задаче одновременного картирования и локализацииRobustness Analysis of Visual SLAM Algorithms with Neural Implicit Map RepresentationАнтиповВ. А.AntipovV. A.

В. А. Антипов, аспирант г.

Санкт-Петербург

V. A. Antipov, Ph.D Student, Engineer

Saint-Petersburg, 191002

v.a.antipov@niuitmo.ruМагазенковЕ. Н.MagazenkovE. N.

Е. Н. Магазенков, магистрант

г. Санкт-Петербург

E. N. Magazenkov, Postgraduate Student

Saint-Petersburg, 191002

enmagazenkov@itmo.ruВедяковА. А.VedyakovA. A.

А. А. Ведяков, канд. тех. наук, доц.

г. Санкт-Петербург

A. A. Vedyakov, Сand. Tech. Sc, Associate Professor

Saint-Petersburg, 191002

vedyakov@itmo.ruУниверситет ИТМОРоссияITMO UniversityRussian Federation2026060220262725965Copyright © Commercial Publisher «New Technologies», 20262026Commercial Publisher «New Technologies»Commercial Publisher «New Technologies»https://mech.novtex.ru/jour/about/submissions#copyrightNoticehttps://mech.novtex.ru/jour/article/view/1920

Работа посвящена анализу робастности к внешним и внутренним возмущениям неявных представлений карт в задаче одновременного картирования и локализации (SLAM) на основе изображений с камеры глубины (RGB-D). На примере известного метода Co-SLAM исследуется чувствительность к трем типам реалистичных возмущений: локальным бликам и засветам, погрешностям калибровки внутренних параметров камеры и аддитивному шуму в канале глубины. Для количественной оценки робастности используется среднеквадратическое отклонение оценки положения камеры после выравнивания по алгоритму Кабша—Умеямы. Результаты моделирования с возмущениями показали высокую робастность к локальным световым эффектам, умеренную — к малым ошибкам калибровки и низкую — к шумам в канале глубины. В работе также предлагается новый метод инициализации начальных положений и ориентации камеры на основании линейных и угловых скоростей, рассчитанных из алгоритма визуальной/визуальноинерциальной одометрии, что обеспечивает повышение точности локализации при зашумленном канале глубины без существенного роста вычислительной сложности.

The study analyzes the robustness of implicit maps to external and internal disturbances in the task of simultaneous localization and mapping (SLAM) based on images from a depth camera (RGB-D). Using Co-SLAM as a representative method, we examined the sensitivity to three realistic perturbations: local bright spots and glare, inaccuracies in the calibration of camera intrinsics, and additive noise in the depth channel. For quantitative robustness evaluation, the root mean square error of the camera position is used after data alignment using the Kabsch-Umeyama algorithm. The results of the simulations with disturbances showed high robustness to local light flares, moderate to small inaccuracies in camera calibration, and low to noise in the depth channel. This study also proposes a new method for determining the initial position and orientation of the camera based on the linear and angular velocities from visual/visual-inertial odometry algorithm, which provides increased accuracy of the localization in the situation with noisy depth channel without a significant increase in computational complexity.

одновременная локализация и построение карты (SLAM)неявное представление картNeRFреконструкция сценыкалибровка камерывизуально-инерциальная одометрияsimultaneous localization and mapping (SLAM)implicit map representationNeRFscene reconstructioncamera calibrationvisual-inertial odometryСтатья подготовлена при финансовой поддержке Министерства науки и высшего образования Российской Федерации, проект № FSER-2025-0002.The article was supported by the Ministry of Science and Higher Education of the Russian Federation (project no. FSER-2025-0002).ReferencesDeTone D., Malisiewicz T., Rabinovich А. SuperPoint: Self-Supervised Interest Point Detection and Description, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 2018, pp. 337—33712, DOI: 10.1109/CVPRW.2018.00060DeTone D., Malisiewicz T., Rabinovich А. SuperPoint: Self-Supervised Interest Point Detection and Description, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 2018, pp. 337—33712, DOI: 10.1109/CVPRW.2018.00060Mildenhall B., Srinivasan P. P., Tancik M., Barron J. T., Ramamoorthi R., Ng R. NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis, Computer Vision — ECCV 2020, 2020, pp. 405—421, DOI: 10.1007/978-3-030-58452-8_24Mildenhall B., Srinivasan P. P., Tancik M., Barron J. T., Ramamoorthi R., Ng R. NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis, Computer Vision — ECCV 2020, 2020, pp. 405—421, DOI: 10.1007/978-3-030-58452-8_24Sucar E., Liu S., Ortiz J., Davison A. J. iMAP: Implicit Mapping and Positioning in Real-Time, Proceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 6229—6238Sucar E., Liu S., Ortiz J., Davison A. J. iMAP: Implicit Mapping and Positioning in Real-Time, Proceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 6229—6238Zhu Z., Peng S., Larsson V., Xu W., Bao H., Cui Z., Oswald M. R., Pollefeys M. NICE-SLAM: Neural Implicit Scalable Encoding for SLAM, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022, pp. 12776—12786, DOI: 10.1109/CVPR52688.2022.01245Zhu Z., Peng S., Larsson V., Xu W., Bao H., Cui Z., Oswald M. R., Pollefeys M. NICE-SLAM: Neural Implicit Scalable Encoding for SLAM, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022, pp. 12776—12786, DOI: 10.1109/CVPR52688.2022.01245Wang H., Wang J., Agapito L. Co-SLAM: Joint Coordinate and Sparse Parametric Encodings for Neural Real-Time SLAM, Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 13293—13302Wang H., Wang J., Agapito L. Co-SLAM: Joint Coordinate and Sparse Parametric Encodings for Neural Real-Time SLAM, Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 13293—13302Zhu S., Wang G., Blum H., Liu J., Song L., Pollefeys M., Wang H. SNI-SLAM: Semantic Neural Implicit SLAM, 2024 IEEE/ CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024, pp. 21167—21177, DOI: 10.1109/CVPR52733.2024.02000Zhu S., Wang G., Blum H., Liu J., Song L., Pollefeys M., Wang H. SNI-SLAM: Semantic Neural Implicit SLAM, 2024 IEEE/ CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024, pp. 21167—21177, DOI: 10.1109/CVPR52733.2024.02000Kong X., Liu S., Taher M., Davison A. J. vMAP: Vectorised Object Mapping for Neural Field SLAM, 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023, pp. 952—961, DOI: 10.1109/CVPR52729.2023.00098Kong X., Liu S., Taher M., Davison A. J. vMAP: Vectorised Object Mapping for Neural Field SLAM, 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023, pp. 952—961, DOI: 10.1109/CVPR52729.2023.00098Shi Y., Yang R., Wu Z., Li P., Liu C., Zhao H., Zhou G. City-scale continual neural semantic mapping with three-layer sampling and panoptic representation, Knowledge-Based Systems, 2024, vol. 284, pp. 111145, DOI: 10.1016/j.knosys.2023.111145Shi Y., Yang R., Wu Z., Li P., Liu C., Zhao H., Zhou G. City-scale continual neural semantic mapping with three-layer sampling and panoptic representation, Knowledge-Based Systems, 2024, vol. 284, pp. 111145, DOI: 10.1016/j.knosys.2023.111145Rosinol A., Leonard J. J., Carlone L. NeRF-SLAM: Real-Time Dense Monocular SLAM with Neural Radiance Fields, 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2023, pp. 3437—3444, DOI: 10.1109/IROS55552.2023.10341922Rosinol A., Leonard J. J., Carlone L. NeRF-SLAM: Real-Time Dense Monocular SLAM with Neural Radiance Fields, 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2023, pp. 3437—3444, DOI: 10.1109/IROS55552.2023.10341922Umeyama S. Least-squares estimation of transformation parameters between two point patterns, IEEE Transactions on Pattern Analysis and Machine Intelligence, 1991, vol. 13 (4), pp. 376—380, DOI: 10.1109/34.88573Umeyama S. Least-squares estimation of transformation parameters between two point patterns, IEEE Transactions on Pattern Analysis and Machine Intelligence, 1991, vol. 13 (4), pp. 376—380, DOI: 10.1109/34.88573Li Q., Ma L., Xu H., Cao M. Investigating SLAM Algorithms for Robustness in Illumination-Varying Scenarios, 2024 6th International Conference on Robotics, Intelligent Control and Artificial Intelligence (RICAI), 2024, pp. 903—907, DOI: 10.1109/RICAI64321.2024.10911790Li Q., Ma L., Xu H., Cao M. Investigating SLAM Algorithms for Robustness in Illumination-Varying Scenarios, 2024 6th International Conference on Robotics, Intelligent Control and Artificial Intelligence (RICAI), 2024, pp. 903—907, DOI: 10.1109/RICAI64321.2024.10911790Muratbekova M., Toganas N., Igali A., Shagyrov M., Kadyrgali E., Yerkin A., Shamoi P. Color Models in Image Processing: А Review and Experimental Comparison, 2025, DOI: 10.48550/ARXIV.2510.00584Muratbekova M., Toganas N., Igali A., Shagyrov M., Kadyrgali E., Yerkin A., Shamoi P. Color Models in Image Processing: А Review and Experimental Comparison, 2025, DOI: 10.48550/ARXIV.2510.00584Sturm J., Engelhard N., Endres F., Burgard W., Cremers D. А benchmark for the evaluation of RGB-D SLAM systems, 2012 IEEE/RSJ International Conference on Intelligent Robots and Systems, 2012, pp. 573—580, DOI: 10.1109/IROS.2012.6385773Sturm J., Engelhard N., Endres F., Burgard W., Cremers D. А benchmark for the evaluation of RGB-D SLAM systems, 2012 IEEE/RSJ International Conference on Intelligent Robots and Systems, 2012, pp. 573—580, DOI: 10.1109/IROS.2012.6385773Taketomi T., Heikkila J. Focal length change compensation for monocular slam, 2015 IEEE International Conference on Image Processing (ICIP), 2015, pp. 4982—4986, DOI: 10.1109/ICIP.2015.7351755Taketomi T., Heikkila J. Focal length change compensation for monocular slam, 2015 IEEE International Conference on Image Processing (ICIP), 2015, pp. 4982—4986, DOI: 10.1109/ICIP.2015.7351755Afzal Maken F., Muthu S., Nguyen C., Sun C., Tong J., Wang S., Tsuchida R., Howard D., Dunstall S., Petersson L. Improving 3D Reconstruction Through RGB-D Sensor Noise Modeling, Sensors, 2025, vol. 25 (3), pp. 950, DOI: 10.3390/s25030950Afzal Maken F., Muthu S., Nguyen C., Sun C., Tong J., Wang S., Tsuchida R., Howard D., Dunstall S., Petersson L. Improving 3D Reconstruction Through RGB-D Sensor Noise Modeling, Sensors, 2025, vol. 25 (3), pp. 950, DOI: 10.3390/s25030950Yen-Chen L., Florence P., Barron J. T., Rodriguez A., Isola P., Lin T.-Y. iNeRF: Inverting Neural Radiance Fields for Pose Estimation, 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2021, pp. 1323—1330, DOI: 10.1109/IROS51168.2021.9636708Yen-Chen L., Florence P., Barron J. T., Rodriguez A., Isola P., Lin T.-Y. iNeRF: Inverting Neural Radiance Fields for Pose Estimation, 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2021, pp. 1323—1330, DOI: 10.1109/IROS51168.2021.9636708Chung C.-M., Tseng Y.-C., Hsu Y.-C., Shi X.-Q., Hua Y.-H., Yeh J.-F., Chen W.-C., Chen Y.-T., Hsu W. H. Orbeez-SLAM: А Real-time Monocular Visual SLAM with ORB Features and NeRF-realized Mapping, 2023 IEEE International Conference on Robotics and Automation (ICRA), 2023, pp. 9400—9406, DOI: 10.1109/ICRA48891.2023.10160950Chung C.-M., Tseng Y.-C., Hsu Y.-C., Shi X.-Q., Hua Y.-H., Yeh J.-F., Chen W.-C., Chen Y.-T., Hsu W. H. Orbeez-SLAM: А Real-time Monocular Visual SLAM with ORB Features and NeRF-realized Mapping, 2023 IEEE International Conference on Robotics and Automation (ICRA), 2023, pp. 9400—9406, DOI: 10.1109/ICRA48891.2023.10160950Geneva P., Eckenhoff K., Lee W., Yang Y., Huang G. OpenVINS: А Research Platform for Visual-Inertial Estimation, 2020 IEEE International Conference on Robotics and Automation (ICRA), 2020, pp. 4666—4672, DOI: 10.1109/ICRA40945.2020.9196524Geneva P., Eckenhoff K., Lee W., Yang Y., Huang G. OpenVINS: А Research Platform for Visual-Inertial Estimation, 2020 IEEE International Conference on Robotics and Automation (ICRA), 2020, pp. 4666—4672, DOI: 10.1109/ICRA40945.2020.9196524Delmerico J., Scaramuzza D. А Benchmark Comparison of Monocular Visual-Inertial Odometry Algorithms for Flying Robots, 2018 IEEE International Conference on Robotics and Automation (ICRA), 2018, pp. 2502—2509, DOI: 10.1109/ICRA.2018.8460664Delmerico J., Scaramuzza D. А Benchmark Comparison of Monocular Visual-Inertial Odometry Algorithms for Flying Robots, 2018 IEEE International Conference on Robotics and Automation (ICRA), 2018, pp. 2502—2509, DOI: 10.1109/ICRA.2018.8460664

The authors declare that there are no conflicts of interest present.