Abstract.

In ADAS (advanced driving assistance system) it is extremely important to be able to identify various weather conditions, especially conditions with low visibility. In this paper, we consider a real-time system for foggy road scenes detection in a video stream from a monocular camera. The system uses an ensemble (committee) of several basic classifiers. A basic classifier is an algorithm based on computer vision which allows you to divide road scenes into several classes of weather conditions, such as rain, snow, fog. Each of the basic classifiers of the system operates with a unique feature space, which makes the system essentially more rigid. The system is a part of the ADAS complex and is used to recommend the speed regime in conditions of reduced visibility. The article presents the results of the experiments.

Keywords:

fog detection, weather conditions recognition, ADAS systems, random forest, committee method, intensity histogram, Laplace operator, discrete Fourier transform, JPEG compression.

PP. 70-77.

Reference

1. Prun V.E., Postnikov V.V., Sadekov R.N., Sholomov D.L. “Development of Active Safety Software of Road Freight Transport, Aimed at Improving Inter-City Road Safety, Based on Stereo Vision Technologies and Road Scene Analysis” // Proceedings of the Scientific-Practical Conference "Research and Development - 2016", Springer, Cham, pp.209-218. – 2017. ISBN 978-3-319-62869-1
2. Bronte, S.; Bergasa, L.M.; Alcantarilla, P.F. “Fog Detection System Based on Computer Vision Techniques”
3. Nicolas Hautière, Jean-Philippe Tarel, Jean Lavenant, and Didier Aubert. “Automatic fog detection and estimation of visibility distance through use of an onboard camera” // Machine Vision Appl., 17(1):8–20, 2006.
4. Middleton W.E.K. “Vision through the Atmosphere” // University of Toronto Press, 1952.
5. Pavlić Mario, et al. "Image based fog detection in vehicles" // Intelligent Vehicles Symposium (IV), 2012 IEEE. IEEE, 2012.
6. Hautière N., Labayrade R., and Aubert D. “Estimation of the visibility distance by stereovision: A generic approach” // IEICE Transactions on information and systems, 89(7):2084–2091, 2006.
7. Caraffa, L. and Tarel, J.-P., “Stereo reconstruction and contrast restoration in daytime fog.” // Proceedings of Asian Conference on Computer Vision (ACCV’12), Part. IV, LNCS, Vol. 7727, Springer, Daejeon, Korea, pp. 13–25., 2013.
8. Mazurov V. D. “ Committee method in the problems of optimization and classification” // M.: Nauka, 1990. — 248 p.
9. Mazurov V. D., Khachai M. Yu. “ Committee constructions” // Izvestiya of Ural State University. 1999. № 14., 1999.
10. Breiman, L. "Random Forests Machine Learning” vol.45 pp.5–32, 2001.
11. Breiman, L. “Out-of-bag estimation” // Technical report, Dept. of Statistics, University of California, Berkeley, 1996.
12. Kurnikov P.A. “Training the decision trees for neural network response validation” // Proceedings of the Seventh International conference «System analysis and information technologies » (SAIT 2017), pp. 256-261, Svetlogorsk, Russia, 2017
13. Prun V., Polevoy D., Postnikov V., "Forward rectification: spatial image normalization for a video from a forward facing vehicle camera," //Proc. SPIE 10341, Ninth International Conference on Machine Vision (ICMV 2016)
14. Prun V., Bocharov D., Koptelov I., Sholomov D., Postnikov V. “Geometric filtration of classification-based object detectors in realtime road scene recognition systems” // Eighth International Conference on Machine Vision (ICMV 2015), Vol. 9875. 98750O, 2015, doi: 10.1117/12.2228709