sibsutisВестник СибГУТИThe Herald of the Siberian State University of Telecommunications and Information Science1998-6920СибГУТИ10.55648/1998-6920-2024-18-4-43-51sibsutis-883Research ArticleСтатьиАлгоритм для автоматического удаления статичных погодных явлений на основе билатерального фильтраAlgorithm for automatic removal of static weather phenomena based on a bilateral filterhttps://orcid.org/0000-0001-5935-9710ЭдельГ. Е.EdelG. E.

Эдель Герман Евгеньевич - аспирант кафедры телевидения и управления, инженер лаборатории телевидения и автоматики,

634050, г. Томск, пр. Ленина, 40.

German E. Edel - Postgraduate Student of the Department of Television and Control, Engineer of the Laboratory of Television and Automation,

40, Lenin Ave., Tomsk, 634050.

german.edel99@gmail.comhttps://orcid.org/0009-0002-7703-3081СукотноваМ. Е.SukotnovaM. E.

Сукотнова Марина Евгеньевна - магистр, инженер лаборатории телевидения и автоматики,

634050, г. Томск, пр. Ленина, 40.

Marina E. Sukotnova - Master, Engineer of the Laboratory of Television and Automation,

40, Lenin Ave., Tomsk, 634050.

msukotnova@mail.ruТомский государственный университет систем управления и радиоэлектроники (ТУСУР)РоссияTomsk State University of Control Systems and Radioelectronics (TSUCSR)Russian Federation202422052024184Вестник СибГУТИ4351Copyright © Эдель Г.Е., Сукотнова М.Е., 20242024Эдель Г.Е., Сукотнова М.Е.Edel G.E., Sukotnova M.E.Данная работа распространяется под лицензией Creative Commons Attribution 4.0.This work is licensed under a Creative Commons Attribution 4.0 License.https://vestnik.sibsutis.ru/jour/article/view/883

Данная статья посвящена предобработке изображений, полученных в условиях плохой видимости. Такие явления, как туман или дымка, способны существенно снижать точность нейронных сетей, предназначенных для обнаружения объектов, тем самым не давая системам, основанным на компьютерном зрении, нормально функционировать. В данной работе был реализован алгоритм для автоматического удаления статичных погодных явлений на основе билатерального фильтра. Алгоритм был протестирован совместно с нейронной сетью, обученной распознавать дорожные знаки.

This article is devoted to the preprocessing of images obtained in poor visibility conditions. Phenomena such as fog or haze can significantly reduce the accuracy of neural networks designed to detect objects. Thus, preventing systems based on computer vision from functioning normally. In this work, an algorithm was implemented for automatic removal of static weather phenomena based on a bilateral filter. The algorithm was tested on a neural network trained to recognize road signs.

компьютерное зрениепредобработка изображенийобнаружение объектовстатичные метеоусловиябилатеральный фильтрповышение точности обнаруженияусловия плохой видимостиcomputer visionimage preprocessingobject detectionstatic weather conditionsbilateral filterincreasing detection accuracypoor visibility conditionsReferencesMathias М., Timofte R., Benenson R. and Van Gool L. Traffic sign recognition — How far are we from the solution?// IEEE International Joint Conference on Neural Networks, 2013. P. 1–8.Mathias М., Timofte R., Benenson R. and Van Gool L. Traffic sign recognition — How far are we from the solution?// IEEE International Joint Conference on Neural Networks, 2013. P. 1–8.Levinson J., Askelad J., Becker J. and Dolson J. Towards fully autonomous driving: Systems and algorithms // IEEE Intelligent Vehicles Symposium (IV), 2011. P. 163–168.Levinson J., Askelad J., Becker J. and Dolson J. Towards fully autonomous driving: Systems and algorithms // IEEE Intelligent Vehicles Symposium (IV), 2011. P. 163–168.Timofte R., Zimmermann K., Van Gool L. Multi-view traffic sign detection, recognition, and 3D localization, Machine vision and applications // Machine vision and applications, 2009. P. 633–647.Timofte R., Zimmermann K., Van Gool L. Multi-view traffic sign detection, recognition, and 3D localization, Machine vision and applications // Machine vision and applications, 2009. P. 633–647.Sermanet P., LeCun Y. Traffic sign recognition with multi-scale convolutional networks // IEEE International Joint Conference on Neural Networks, 2011. P. 2809–2813.Sermanet P., LeCun Y. Traffic sign recognition with multi-scale convolutional networks // IEEE International Joint Conference on Neural Networks, 2011. P. 2809–2813.Houben S., Stallkamp J., Salmen J., Schlipsing M. and Igel C. Detection of traffic signs in real-world images: The German Traffic Sign Detection Benchmark // IEEE International Joint Conference on Neural Networks, 2013. P. 1–8.Houben S., Stallkamp J., Salmen J., Schlipsing M. and Igel C. Detection of traffic signs in real-world images: The German Traffic Sign Detection Benchmark // IEEE International Joint Conference on Neural Networks, 2013. P. 1–8.Qi Y., Zhang P. The Tracker with Online Training Based on the TLD Algorithm // Proceedings of the 2nd International Conference on Information Technology and Intelligent Transportation Systems, 2017. P. 76–81.Qi Y., Zhang P. The Tracker with Online Training Based on the TLD Algorithm // Proceedings of the 2nd International Conference on Information Technology and Intelligent Transportation Systems, 2017. P. 76–81.Dang P., Tran N. T., To V. H. and Thi M. T. Improved YOLOv5 for real-time traffic signs recognition in bad weather conditions // The Journal of Supercomputing, vol. 79, no. 13, 2023. P. 10706–10724.Dang P., Tran N. T., To V. H. and Thi M. T. Improved YOLOv5 for real-time traffic signs recognition in bad weather conditions // The Journal of Supercomputing, vol. 79, no. 13, 2023. P. 10706–10724.Xing J., Nguyen M. and Yan W. Q. The Improved Framework for Traffic Sign Recognition Using Guided Image Filtering // SN Computer Science, 2022. P. 85–99.Xing J., Nguyen M. and Yan W. Q. The Improved Framework for Traffic Sign Recognition Using Guided Image Filtering // SN Computer Science, 2022. P. 85–99.

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