Hydrophysics

Hydrophysics

Automatic Rain Detection from Digital Videos Using the Farneback Optical Flow AlgorithmAutomatic Rain Detection from Digital Videos Using the Farneback Optical Flow Algorithm

Document Type : Original Article

Authors
Fateme Heydari 1
Asghar Milan 2
Alireza Vafaeinejad 1
1 Faculty of Civil, Water and Environmental, Shahid Beheshti University, Tehran, Iran.
2 Faculty of Civil, Water and Environmental, Engineering Shahid Beheshti university, Tehran, Iran.
Abstract
Automatic rain detection from digital video, especially in regions lacking traditional instrumentation, provides a low-cost and effective tool for atmospheric monitoring and early warning. In this study, the Farneback optical flow algorithm was employed to identify rain-induced motion in consecutive video frames. By analyzing the predominantly vertical motion of pixels, the method can detect rainfall and functions as a simple and relatively robust approach under real-world conditions. A set of rainy and non-rainy videos was used for evaluation, and the results were compared against manually labeled reference data. The algorithm achieved an average accuracy of 78.72%, demonstrating its adequate capability for rain detection, particularly in videos with stable frame rates and relatively high illumination. The findings indicate that this method can yield improved performance when combined with other techniques and further training, while remaining cost-effective and straightforward to implement. The approach is well suited to real-time and surveillance applications and can act as a low-cost alternative to conventional monitoring equipment. Given the importance of rain detection in natural disaster management and climate change studies, this field has become a critical area of recent research, and future enhancements are expected to further increase efficiency and accuracy.
Keywords
Automatic detection of raindrops and rain streaks
image processing
Farneback optical flow algorithm
CCTV videos
Subjects
[1] Rahayu SA, Sipayung SB, Witono A, Suprihatin LS. The Performance of K-Nearest Neighbor (KNN) Approach for Estimating Extreme Rain Events based on CCTV Images Camera Data. Jurnal EECCIS (Electrics, Electronics, Communications, Controls, Informatics, Systems). 2023 Sep 1;17(2):59-65.
[2] Hakim AL, Dewi R. Automatic rain detection system based on digital images of CCTV cameras using convolutional neural network method. InIOP Conference Series: Earth and Environmental Science 2021 Nov 1 (Vol. 893, No. 1, p. 012048). IOP Publishing.
[3] Fiallos-Salguero M, Khu ST, Guan J, Wang T, Wang M. Computer Vision-Based Method for Rainfall Estimation Using CCTV Cameras and Smartphone Videos. In15th International Conference on Hydroinformatics 2024 (p. 284).
[4] Lee J, Byun J, Baik J, Jun C, Kim HJ. Estimation of raindrop size distribution and rain rate with infrared surveillance camera in dark conditions. Atmospheric Measurement Techniques. 2023 Feb 8;16(3):707-25.
[5] Farnebäck G. Two-frame motion estimation based on polynomial expansion. InScandinavian conference on Image analysis 2003 Jun 24 (pp. 363-370). Berlin, Heidelberg: Springer Berlin Heidelberg.
[6] Tripathi AK, Mukhopadhyay S. Removal of rain from videos: a review. Signal, Image and Video Processing. 2014 Nov;8(8):1421-30.
[7] Garg K, Nayar SK. Vision and rain. International Journal of Computer Vision. 2007 Oct;75(1):3-27.
[8] Barnum PC, Narasimhan S, Kanade T. Analysis of rain and snow in frequency space. International journal of computer vision. 2010 Jan;86(2):256-74.
[9] Roser M, Geiger A. Video-based raindrop detection for improved image registration. In2009 IEEE 12th International Conference on Computer Vision Workshops, ICCV Workshops 2009 Sep 27 (pp. 570-577). IEEE.
[10] Haurum JB, Bahnsen CH, Moeslund TB. Is it Raining Outside? Detection of Rainfall using General-Purpose Surveillance Cameras. InCVPR Workshops 2019 Aug 12 (pp. 55-64).
[11] Shorman SM, Pitchay SA. A review of rain streaks detection and removal techniques for outdoor single image. ARPN J. Eng. Appl. Sci. 2016 May;11(10):6303-8.
[12] Wang H, Xie Q, Zhao Q, Meng D. A model-driven deep neural network for single image rain removal. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition 2020 (pp. 3103-3112).
[13] Yang W, Tan RT, Wang S, Liu J. Self-learning video rain streak removal: When cyclic consistency meets temporal correspondence. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition 2020 (pp. 1720-1729).
[14] Yan W, Tan RT, Yang W, Dai D. Self-aligned video deraining with transmission-depth consistency. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition 2021 (pp. 11966-11976).  
[15] Islam MR, Paul M. Video rain-streaks removal by combining data-driven and feature-based models. Sensors. 2021 Oct 15;21(20):6856.
[16] Yi Z, Shi H, Jiang Q, Gao Y, Wang Z, Zhang Y, Yang K, Wang K. Benchmarking the Robustness of Optical Flow Estimation to Corruptions. arXiv preprint arXiv:2411.14865. 2024 Nov 22.
[17] Gonzalez, R.C. and Woods, R.E. (2008) Digital Image Processing. 3rd Edition, Pearson Prentice Hall, Upper Saddle River. - References - Scientific Research Publishing, www.scirp.org. https://www.scirp.org/reference/ReferencesPapers?ReferenceID=1258015
[18] Stauffer C, Grimson WE. Adaptive background mixture models for real-time tracking. InProceedings. 1999 IEEE computer society conference on computer vision and pattern recognition (Cat. No PR00149) 1999 Jun 23 (Vol. 2, pp. 246-252). IEEE.
[19] Zivkovic Z. Improved adaptive Gaussian mixture model for background subtraction. InProceedings of the 17th International Conference on Pattern Recognition, 2004. ICPR 2004. 2004 Aug 26 (Vol. 2, pp. 28-31). IEEE.
[20] Deng LJ, Huang TZ, Zhao XL, Jiang TX. A directional global sparse model for single image rain removal. Applied Mathematical Modelling. 2018 Jul 1;59:662-79.
[21] Navin, M.J.R. and Pankaja, R. Performance analysis of text classification algorithms using confusion matrix. International Journal of Engineering and Technical Research (IJETR). 2016; 869–2321.
[22] Sokolova M, Lapalme G. A systematic analysis of performance measures for classification tasks. Information processing & management. 2009 Jul 1;45(4):427-37.
[23] Li H, Luo K, Zeng B, Liu S. Gyroflow+: Gyroscope-guided unsupervised deep homography and optical flow learning. International Journal of Computer Vision. 2024 Jun;132(6):2331-49.
[24] Li Z, Hyeon J, Choi HJ. Rainfall recognition from road surveillance videos using TSN. Journal of Korean Society for Atmospheric Environment. 2018;34(5):735-47.
 

  • Receive Date 14 December 2025
  • Revise Date 31 December 2025
  • Accept Date 06 February 2026