 |
O. S. Makarov, E. V. Shchennikova Numerical Method for Recognizing Shadows of Moving Objects |
 |
|
Abstract.
Motion detection of objects in video is a computationally intensive area of computer vision that is prone to numerous errors. Illumination changes can cause shadows to appear in the video, which are often misinterpreted as part of a moving object or even detected as a separate object. This paper proposes a new numerical method for detecting and removing shadows of moving objects in the context of motion recognition in video recordings. The relevance of this work stems from the lack of universal standalone methods for shadow recognition that are not integrated into specific motion detection algorithms. The proposed approach takes into account spatial correlations between neighboring pixels combined with the normal distribution of noise. Testing has shown that the application of this method significantly improves background extraction results with only a minor increase in the computational complexity of the recognition algorithm, allowing its use in real-time and resource-constrained environments.
Keywords:
motion detection, background subtraction, numerical method, shadow detection, foreground mask, background model, optimization, normal distribution.
DOI 10.14357/20718632260105
EDN KWAXJJ
PP. 54-65.
References
1. Valanarasu JMJ, Patel VM. Fine-Context Shadow Detection Using Shadow Removal. 2023 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV). 2023:1-10. https://doi.org/10.1109/WACV56688.2023.00175. 2. Abdusalomov A, Whangbo TK. Detection and Removal of Moving Object Shadows Using Geometry and Color Information for Indoor Video Streams. Applied Sciences. 2019;9(23):5165. https://doi.org/10.3390/app9235165 3. Zhou D, Liu Y, Li K. Moving Object Detection Based on Ghost and Shadow Removal. Journal of Physics: Conference Series. The 4th International Conference on Modeling, Simulation, Optimization and Algorithm (ICMSOA 2022). 2022; 2508:012035. https://doi.org/10.1088/1742-6596/2508/1/012035 4. Barbuzza R, Dominguez L, Pérez A, Esteberena L, Rubiales A, D'Amato J. A Shadow Removal Approach for a Background Subtraction Algorithm. Communications in Computer and Information Science. 2018. https://doi.org/10.1007/978-3-319- 75214-3_10 5. Agrawal S, Natu P. ABGS Segmenter: Pixel Wise Adaptive Background Subtraction and Intensity Ratio Based Shadow Removal Approach for Moving Object Detection. The Journal of Supercomputing. 2022;79:3628-3670. https://doi.org/10.1007/s11227-022 -04972-9 6. Varghese A, G S. Sample-Based Integrated Background Subtraction and Shadow Detection. IPSJ Transactions on Computer Vision and Applications. 2017;9(25). https://doi.org/10.1186/s41074-017-0036-1 7. ChangeDetection.NET (CDNET). A video database for testing change detection algorithms. Available from: http://changedetection.net/ [Accessed 18 May 2025]. 8. Makarov O.S., Shchennikova E.V. Analysis of background subtraction algorithms. In: International Conference on Business Economics, Management, Engineering Technology, Medical and Health Sciences. 2021, Morrisville, USA. p. 65–77. (In Russ). 9. Makarov O.S. Statistical modeling of boundaries in object motion recognition on video recordings. Inzhenerny vestnik Dona. 2023. (In Russ). Available from: https://www.ivdon.ru/back_media/uploads/article/pdf/IVD_29__11_ makarov.pdf_47a8594701.pdf [Accessed 27 July 2025].
|
