Abstract
The main goal of the presented work is to optimize the developed intelligent system for recognizing and detecting vehicles on video data using the YOLOv3 convolutional neural network. Basic results are obtained for real work conditions with the use of graphic processors. In addition, a special performance study was made for the Intel Core i5–8500 CPU. Optimization is based not only on classical neural network methods, such as model pruning, but also modified procedures have been proposed for efficient processing of video information, in particular, optical flow and motion prediction. After the optimization, the data processing speed increased by 4 times when using the NVIDIA RTX 2080 Super GPU and amounted to about 30 frames per second. CPU acceleration was achieved using the Intel OpenVINO toolkit. Performance on the CPU reached almost the same values as on the video card, and the acceleration was almost 30 times from slowest model to fastest. It is important to note that the implementation of optical flow and motion extrapolation was not required on the CPU.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Andriyanov, N., Dementiev, V., Kondratiev, D.: Tracking of Objects in Video Sequences. In: Czarnowski, I., Howlett, R.J., Jain, L.C. (eds.) Intelligent Decision Technologies. SIST, vol. 238, pp. 253–262. Springer, Singapore (2021). https://doi.org/10.1007/978-981-16-2765-1_21
Li, G., Zhang, D., Zeng, J., Chen, S.: Vehicle monitor system for public transport management based on embedded technology. Phys. Procedia 24(2), 953–960 (2012). https://doi.org/10.1016/j.phpro.2012.02.143
Khan, S.U., Alam, N., Jan, S.U., Koo, I.S.: IoT-Enabled vehicle speed monitoring system. Electronics 11, 614 (2022). https://doi.org/10.3390/electronics11040614
Cai, Z., Vasconcelos N.: Cascade R-CNN: High Quality Object Detection and Instance Segmentation, https://arxiv.org/abs/1906.09756, Accessed 27 Aug 2022
Carion, N., Massa, F., Synnaeve, G., Usunier, N., Kirillov, A., Zagoruyko, S.: End-to-End Object Detection with Transformers, https://arxiv.org/abs/2005.12872, Accessed 27 Aug 2022
Song, X., Gu, W.: Multi-objective real-time vehicle detection method based on yolov5. In: 2021 International Symposium on Artificial Intelligence and its Application on Media (ISAIAM) Proceedings, pp. 142–145 (2021). https://doi.org/10.1109/ISAIAM53259.2021.00037
Andriyanov, N.: Analysis of the acceleration of neural networks inference on Intel processors based on OpenVINO Toolkit. In: Proceedings of 2020 Systems of Signal Synchronization, Generating and Processing in Telecommunications, SYNCHROINFO, pp. 1–6 (2020). https://doi.org/10.1109/SYNCHROINFO49631.2020.9166067
Arnautović, A., Teskeredzic, E.: Evaluation of artificial neural network inference speed and energy consumption on embedded systems. INFOTEH 1, 1–5 (2021). https://doi.org/10.1109/INFOTEH51037.2021.9400658
Zhang, X., Zou, J., He, K., Sun, J.: Accelerating very deep convolutional networks for classification and detection, https://arxiv.org/pdf/1505.06798.pdf, Accessed 28 Aug 2022
Xiao, B., Shi, W., Lu, G.: An optimized quantization technique for image compression using discrete tchebichef transform. Pattern Recognit. Image Anal. 28, 371–378 (2018). https://doi.org/10.1134/S1054661818030021
Novac, P.E., Boukli, G.H., Pegatoquet, A., Miramond, B., Gripon, V.: Quantization and deployment of deep neural networks on microcontrollers. Sensors 21, 2984 (2021). https://doi.org/10.3390/s21092984
Shirokanev, A.S., Andriyanov, N.A., Ilyasova, N.Y.: Development of vector algorithm using CUDA technology for three-dimensional retinal laser coagulation process modeling. Comput. Opt. 45(3), 427–437 (2021). https://doi.org/10.18287/2412-6179-CO-828
Andriyanov, N.A., Vasil'ev, K.K.; Dement'ev, V.E.: Investigation of filtering and objects detection algorithms for a multizone image sequence. International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences - ISPRS Archives 42, is. 2/W12, 7–10 (2019). https://doi.org/10.5194/isprs-archives-XLII-2-W12-7-2019
Borman, S., Stevenson, R.: Image Sequence Processing. University of Notre Dame, Paris (2002)
Anitha, E., Jiji, C.V.: Optical acceleration for motion description in videos. In: CVPR 2017 Proceedings, pp. 1–9 (2017)
Flynn, J., Neulander, I., Philbin, J., Snavely, N.: Deepstereo: Learning to predict new views from the world’s imagery. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 5515–5524 (2016)
Andriyanov, N.A., Dementiev, V.E., Tashlinskii, A.G.: Detection of objects in the images: from likelihood relationships towards scalable and efficient neural networks. Comput. Opt. 46(1), 139–159 (2022). https://doi.org/10.18287/2412-6179-CO-922
Girshick, R.: Fast R-CNN. In: Proceedings of International Conference on Computer Vision (ICCV), vol. 1, pp. 1440–1448 (2015). https://doi.org/10.1109/ICCV.2015.169
Redmon, J., Farhadi, A.: YOLOv3: An incremental improvement, https://arxiv.org/abs/1804.02767 Accessed 30 Aug 2022
Dementyiev, V.E., Andriyanov, N.A., Vasilyiev, K.K.: Use of images augmentation and implementation of doubly stochastic models for improving accuracy of recognition algorithms based on convolutional neural networks. In: Proceedings of 2020 Systems of Signal Synchronization, Generating and Processing in Telecommunications, SYNCHROINFO, pp. 1–4 (2020). https://doi.org/10.1109/SYNCHROINFO49631.2020.9166000
Buslaev, A., Iglovikov, V.I., Khvedchenya, E., Parinov, A., Druzhinin, M., Kalinin, A.A.: Albumentations: fast and flexible image augmentations. Information 11, 125 (2020). https://doi.org/10.3390/info11020125
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 Springer Nature Switzerland AG
About this paper
Cite this paper
Andriyanov, N., Dementiev, V., Tashlinskiy, A. (2023). Optimization of the Computer Vision System for the Detection of Moving Objects. In: Rousseau, JJ., Kapralos, B. (eds) Pattern Recognition, Computer Vision, and Image Processing. ICPR 2022 International Workshops and Challenges. ICPR 2022. Lecture Notes in Computer Science, vol 13644. Springer, Cham. https://doi.org/10.1007/978-3-031-37742-6_32
Download citation
DOI: https://doi.org/10.1007/978-3-031-37742-6_32
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-37741-9
Online ISBN: 978-3-031-37742-6
eBook Packages: Computer ScienceComputer Science (R0)