Efficient algorithm of transport identification in video surveillance systems

The objective is to enhance transport identification accuracy by stationary and movable video cameras. The poor quality of images or their parts due to some external or internal impacts can lead to losses of important information. The adverse effects to affect the picture quality include the excessive scene brightness, high velocity of transport and insufficient capabilities of cameras to shoot in the dark or in the excessive brightness. The bright sources of light, for instance, automobile headlights, feed the excessive light to the camera matrix and make some parts of images indistinguishable, while high velocity of vehicles produces a kind of tail behind a vehicle in the image when the exposure time is insufficient. As a result, the test subjects in the images become unsuitable for the analysis. For instance, it becomes sometimes impossible to read precisely the figures on the license plate of a car. High diffuseness or lightshorting can make the vehicle registration plates of Russia unreadable and undistinguishable for any identification system. This problem is similarly intrinsic to mining facilities with stationary video surveillance of production processes. This article focuses on automated estimation of the diffuseness or light-shorting rate in the images of the vehicle registration plates. The methods are proposed for the quick estimate of the diffuseness or light-shorting rate as a casestudy of the vehicle registration plates. These methods are suitable for embedding in different systems on the basis of NVIDIA Jetson Nano Developer Kit platform without significant delays in operation. Alongside with distinguishing between the readable and unreadable images, this method enables quantification of diffuseness and light-shorting. The obtained results can assist in adjustment of such camera parameters as exposure and diaphragm, which can improve the quality of snaps.

Epifanov V. A., Temkin I. O., Krasnoyaruzhskiy S. E. Efficient algorithm of transport identification in video surveillance systems. MIAB. Mining Inf. Anal. Bull. 2023;(6):5-18. [In Russ]. DOI: 10.25018/0236_1493_2023_6_0_5.

V.A. Epifanov^1,2, Graduate Student, e-mail: epifanov_vld@yandex.ru,
I.O. Temkin¹, Dr. Sci. (Eng.), Professor, e-mail: igortemkin@yandex.ru,
S.E. Krasnoyaruzhskiy², Developer, e-mail: blazzer4267@gmail.com,
¹ National University of Science and Technology «MISiS», 119049, Moscow, Russia,
² «Citylabs» LLC, Moscow, 121205, Russia.

I.O. Temkin, e-mail: igortemkin@yandex.ru.

1. Arena F., Pau G., Severino A. An overview on the current status and future perspectives of smart cars. Infrastructures. 2020, vol. 5, no. 7, article 53. DOI: 10.3390/infrastructures5070053.

2. Temkin I., Myaskov A., Deryabin S., Konov I., Ivannikov A. Design of a digital 3D model of transport—technological environment of open-pit mines based on the common use of telemetric and geospatial information. Sensors. 2021, vol. 21, no. 18, article 6277. DOI: 10.3390/ s21186277.

3. Temkin I. O., Klebanov D. A., Deryabin S. A., Konov I. S. Construction of intelligent geoinformation system for a mine using forecasting analytics techniques. MIAB. Mining Inf. Anal. Bull. 2020, no. 3, pp. 114—125. [In Russ]. DOI: 10.25018/0236-1493-2020-3-0-114-125.

4. Zhang H., Tao P., Meng X., Liu M., Liu X. An optimum deployment algorithm of camera networks for open-pit mine slope monitoring. Sensors. 2021, vol. 21, no. 4, article 1148. DOI: 10.3390/s21041148.

5. Brooks C. N., Dobson R. J., Dean D. B., Banach D., Oommen T., Havens T. C., Ahlborn T., Cook S. J., Clover A. Evaluating the use of unmanned aerial vehicles for transportation purposes. Report No. RC-1616. Michigan Department of Transportation, Michigan, USA, 2015, 201 р.

6. Bolotova Y. A., Spitsyn V. G., Rudometkina M. N. License plate recognition algorithm on the basis of a connected components method and a hierarchical temporal memory model. Computer optics. 2015, vol. 39, no. 2, pp. 275—280. [In Russ]. DOI: 10.18287/0134-2452-2015-39-2-275-280.

7. Poltavskii A. V., Yurushkina T. G., Yurushkin M. V. Automatic license-plate recognition. Advanced Engineering Research. 2020, vol. 20, no. 1, pp. 93—99. [In Russ]. DOI: 10.23947/19925980-2020-20-1-93-99.

8. Yang X., Wang X. Recognizing license plates in real-time. arXiv: 2019. 1906.04376. DOI: 10.48550/arXiv.1906.04376.

9. Wang J., Hu X. Convolutional neural networks with gated recurrent connections. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2022, vol. 44, pp. 3421—3435. DOI: 10.1109/TPAMI.2021.3054614.

10. Wang J., Hu Y. Gated recurrent convolution neural network for OCR. NIPS'17: Proceedings of the 31st International Conference on Neural Information Processing Systems. 2017, pp. 334—343.

11. Atienza R. Vision transformer for fast and efficient scene text recognition. Document Analysis and Recognition — ICDAR. 2021, pp. 319—334. DOI: 10.1007/978-3-030-86549-8_21.

12. Kissos I., Dershowitz N. OCR error correction using character correction and featurebased word classification. 12th IAPR Workshop on Document Analysis Systems (DAS). 2016. DOI: 10.1109/DAS.2016.44.

13. Mokhtar K., Bukhari S., Dengel A. OCR Error Correction: State-of-the-art vs An NMT Based Approach. 13th IAPR International Workshop on Document Analysis Systems (DAS). 2018. DOI: 10.1109/DAS.2018.63.

14. Jatzkowski I., Wilke D., Maurer M. A deep-learning approach for the detection of overexposure in automotive camera images. 21st International Conference on Intelligent Transportation Systems (ITSC). 2018. DOI: 10.1109/ITSC.2018.8569692.

15. Singh M., Tiwari R., Swami K., Vijayvargiya A. Detection of glare in night photography. 23rd International Conference on Pattern Recognition (ICPR). 2016. DOI: 10.1109/ICPR.2016.7899744.

16. Chen Y., Liu F., Pei K. Self-supervised sun glare detection CNN for self-aware autonomous driving. Machine Learning for Autonomous Driving Workshop at the 35th Conference on Neural Information Processing System. 2021.

17. Zotin A. G., Pakhirka A. I., Damov M. V., Savchina E. I. Visual quality enhancement of the image based on infrared data in video surveillance systems. Software and Systems. 2016, no. 3, pp. 109—120. [In Russ]. DOI: 10.15827/0236-235X.115.109-120.

18. Xin Yang, Ke Xu, Yibing Song, Qiang Zhang, Xiaopeng Wei, Rynson Lau Image correction via deep reciprocating HDR transformation. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). 2018. DOI: 10.1109/CVPR.2018.00193.

19. Zaidi S., Ansari M., Aslam A., Kanwal N., Asghar M., Lee B. A Survey of modern deep learning based object detection models. Digital Signal Processing. 2022, vol. 126. DOI: 10.1016/ j.dsp.2022.103514.

20. Kai Feng, Han Hong, Ke Tang, Jingyuan Wang Decision making with machine learning and ROC curves. SSRN Electronic Journal. 2019/06/01. DOI: 10.2139/ssrn.3382962