Abstract
The purpose of the study is to develop an automated system that improves the accuracy and reliability of the information obtained when identifying the condition of typical objects located on the industrial building roofs. The research is aimed at developing an algorithmic and software module for identifying typical objects. To achieve this goal, the construction of algorithmic software for the identification of typical objects in images of industrial buildings, the synthesis of a software module based on the constructed algorithms, and testing of the software module on test examples were performed. The study was conducted on the territory of the city-forming enterprise of ferrous metallurgy and using the laboratory base of the Federal State Budgetary Educational Institution of Higher Education «Nosov Magnitogorsk State Technical University». It took more than three years to conduct the study, starting in July 2021. The research uses the following methods: monitoring the technical condition of industrial buildings, photo and video shooting, namely, collecting graphic information using a rotary-type unmanned aerial vehicle (UAV), creating an algorithmic and software module and testing it to search for typical objects, measuring the level of detection accuracy in various photographs (photographs with different average brightness values), comparison of image processing results for two versions of the software module (console and Web-versions). Within the framework of the study, two types of typical objects on industrial tasks are considered: ventilation deflectors and wind fences located on the roof of industrial buildings. The software product allows you to track the number, location and technical condition (availability, design position) of typical industrial buildings and structures. As part of the study, testing was performed on test cases for two versions of the module, the console version and the Web-version with a user interface. The results of the testing have been structured. The features of the console and Web-versions of the module are noted. The error of detecting typical objects during inspection by a specialist is excluded, due to the high subjectivity of expert assessment during the inspection of industrial buildings and structures. Automation of the technical condition monitoring for industrial buildings has been carried out.
Keywords
algorithmic support and software, unmanned aerial vehicle, digital image, standard object, testing, software module, identification of the condition of standard objects, images of industrial buildings, ventilation deflectors, wind fences
1. New production facilities in Russia – 2023. Novye proizvodstva Rossii – 2023. Available at: https://tenchat.ru/media/1444564-novyye-proizvodstva-rossii--2023 (accessed 24 February 2024). (In Russian)
2. Industrial safety: briefly about the results of 2023 and plans for 2024. Promyshlennaya bezopasnost: korotko ob itogakh 2023 goda i planakh na 2024 god. Available at: https://portal.edufire37.ru/news/201 (accessed 24 February 2024). (In Russian)
3. Narkevich M.Yu. Razvitie metodologii sozdaniya sistemy menedzhmenta kachestva metallurgicheskogo predpriyatiya, ekspluatiruyushchego opasnye proizvodstvennye objekty, na osnove prikladnoy tsifrovoy platformy. Doct. Diss. [Development of methodology for creating a quality management system for a metallurgical enterprise operating hazardous production facilities based on the applied digital platform. Doct. Diss.]. Magnitogorsk, 2023. 97 p. (in Russian)
4. Narkevich M., Logunova O., Kornienko V., Kalitaev A., Egorova L., Nikolaev A., Tyulyumov A., Zlydarev N. An empirical approach to quality assurance of materials, products and structures in industrial safety. Transportation Research Procedia. 2022, vol. 63, pp. 119-128. doi: 10.1016/j.trpro.2022.05.014
5. Zlydarev N.V., Tyulyumov A.N., Kornienko V.D., Narkevich M.Yu. Software product for assessing defects in industrial structures: search for typical objects. Matematicheskoe i programmnoe obespechenie sistem v promyshlennoy i sotsialnoy sferakh [Mathematical and software systems in the industrial and social spheres], 2022, no. 10(2), pp. 12–16. (in Russian). doi: 10.18503/2306-2053-2022-10-2-12-16
6. Narkevich M.Yu., Logunova O.S., Kalandarov P.I. et al. Results of a pilot experiment on monitoring the condition of buildings and structures using unmanned aerial vehicles. IOP Conference Series: Earth and Environmental Science: 2nd International Conference on Energetics, Civil and Agricultural Engineering 2021 (ICECAE 2021). IOP Science, 2021, vol. 939. 012030. doi: 10.1088/1755-1315/939/1/012030
7. Narkevich M.Yu., Kornienko V.D., Polyakova M.A. Visual control as a basis for the development of automated systems for remote monitoring and assessment of the quality of buildings and structures at hazardous production facilities. Izvestiya Tulskogo Gosudarstvennogo Universiteta. Tekhnicheskie NAUKI [Izvestiya Tula State University. Technical science], 2021, no. 5, pp. 570–576. (in Russian). doi: 10.24412/2071-6168-2021-5-570-576
8. Egorov V.V., Kardashevsky A.P. Telemetricheskoe ustroystvo sbora informatsii i monitoringa za udalennym obyektom [Telemetric device for collecting information and monitoring a remote object]. Patent RF, no. 180206 U1, 2018.
9. Tarasyan V.S., Karachev D.K. Automatic system for searching objects in an image using the example of a license plate. Innovatsionnyy Transport [Innotrans], 2018, no. 2(28), pp. 34-38 (in Russian). doi: 10.20291/2311-164x-2018-2-34-38
10. Milov V.R., Kuvshinov A.S. Programma dlya testirovaniya algoritmov poiska shablonnykh obyektov na izobrazheniyakh [Program for testing algorithms for searching for template objects in images], Patent RF, no. 2018610611, 2018.
11. Shoberg A.G., Mikhailenko P.S. Modul sravneniya izobrazheniy na osnove chastotnykh koeffitsientov diskretnogo kosinusnogo preobrazovaniya dlya sistemy poiska obyektov na snimkakh [Image comparison module based on frequency coefficients of discrete cosine transform for a system for searching objects in images], Patent RF, no. 2017617668, 2017.
12. Patrushev A.O., Osipov M.P. Search for the boundaries of rectangular objects in images with a contour representation format. Fundamentalnye Issledovaniya [Fundamental Research], 2017, no. 12(1), pp. 91-96. (in Russian)
13. Zakharov A.A., Barinov A.E., Zhiznyakov A.L., Titov V.S. Searching for objects in images using a graph-based structural descriptor. Kompyuternaya optika [Computer Optics], 2018, no. 42(2), pp. 283–290. (in Russian). doi: 10.18287/2412-6179-2018-42-2-283-290
14. Kaiyang L., Gang H., Yuanlin Z., Guangfeng L., Congjun C. Approximate object location deep visual representations for image retrieval. Displays. 2023, vol. 77, 102379. doi: 10.1016/j.displa.2023.102376
15. Xian Z., Zhibin Q., Qiang L, Dejun Z., Wankou Y. Searching Enhanced Decoder with switchable skip connection for semantic segmentation. Pattern Recognition. 2024, vol. 149, 110196. doi: 10.1016/j.patcog.2023.110196
16. Wei C., Yan L., Zijian T., Fan Z. 2D and 3D object detection algorithms from images: A Survey. Array. 2023, vol. 19, 100305. doi: 10.1016/j.array.2023.100305
17. Cheng P., Anoop C., Rushi L., Xiaonan L., Hongxun Y. Searching by parts: Towards fine-grained image retrieval respecting species correlation. Gene Expression Patterns. 2023, vol. 47, 119304. doi: 10.1016/j.gep.2023.119304
18. Peimin Y., Shengbo Y., Qiuyu Z., Tao Z., Wennan C. Real-time detection and tracking of infrared small targets based on grid fast density peaks searching and improved KCF. Infrared Physics & Technology. 2022, vol. 123, 104181. doi: 10.1016/j.infrared.2022.104181
19. Zhangxi Y., Kaile Y., Yuwei L., Shijie G., Yiming S., Xunlong C., Riwen L., Houxi Z. A comparison between Pixel-based deep learning and Object-based image analysis (OBIA) for individual detection of cabbage plants based on UAV Visible-light images. Computers and Electronics in Agriculture. 2023, vol. 209, 107822. doi: 10.1016/j.compag.2023.107822
20. Sirota A.A., Mitrofanova E.Yu., Milovanova A.I. Analysis of algorithms for searching objects in images using various modifications of convolutional neural networks. Vestnik Voronezhskogo Gosudarstvennogo Universiteta. Seriya: Sistemnyy Analiz i Informatsionnye Tekhnologii [Proceedings of Voronezh State University. Series: Systems Analysis and Information Technologies], 2019, no. 3, pp. 123-137. (in Russian)
21. Vizilter Yu.V., Gorbatsevich V.S., Vishnyakov B.V., Sidyakin S.V. Searching for objects in an image using morphlet descriptions. Kompyuternaya optika [Computer Optics], 2017, no. 41(3), pp. 406-411. (in Russian). doi: 10.18287/2412-6179-2017-41-3-406-411
22. Butenko V.A. Modul poiska koordinat obyektana izobrazhenii [Module for searching the coordinates of an object in an image], Patent RF, no. 2018663741, 2018.
23. Ponomarev A.V., Bogoslovsky A.V., Zhigulina I.V., Balzhak A.E. Sposob poiska obyektov na tsifrovykh izobrazheniyakh [Method for searching objects in digital images], Patent RF, no. 2718172 C1, 2019.
24. Kulikov S.S. Testirovanie programmnogo obespecheniya, Ch.3 [Software testing, P.3]. Available at: https://svyatoslav.biz/software_testing_book/ (accessed 23 February 2024). (In Russian)
Narkevich M.Yu., Logunova O.S., Zlydarev N.V., Tyulyumov A.N., Kornienko V.D. Algorithmic and software module for typical objects identification in images of industrial buildings. Elektrotekhnicheskie sistemy i kompleksy [Electrotechnical Systems and Complexes], 2024, no. 2(63), pp. 80-89. (In Russian). https://doi.org/10.18503/2311-8318-2024-2(63)-80-89