Görüntü İşleme Tekniklerine Dayanan Tehlikeli Madde Tespit ve Uyarı Yaklaşımı

Abstract

Tehlikeli maddeler, sanayi, lojistik, tarım ve enerji gibi birçok sektörde yaygın olarak kullanılmakla birlikte, potansiyel olarak ciddi riskler taşımaktadır. Bu risklerin, maddeler taşımacılık süreçlerine başlamadan önce doğru bir şekilde belirlenmesi, olası kazaların önlenmesi ve risklerin minimize edilmesi açısından kritik öneme sahiptir. Bu çalışma, tehlikeli madde taşımacılığında meydana gelebilecek kazaların önlenmesi, olası kaza durumlarında hızlı, etkili müdahale sağlanması ve erken önlem alınmasına yönelik bir yaklaşım sunmaktadır. Çalışmada, görüntü işleme tekniklerinden biri olan Optik Karakter Tanıma (Optical Character Recognition-OCR) teknolojisi kullanılmıştır. Tehlikeli madde etiketleri OCR algoritmaları yardımıyla tespit edilmiş ve etiket üzerindeki metinler başarılı bir şekilde algılanmıştır. Algılanan metinler, özellikle tehlikeli maddelere özgü Birleşmiş Milletler (United Nations-UN) numaraları, önceden oluşturulan bir veri tabanı ile eşleştirilmiştir. Veri tabanı ile eşleştirilen UN numaraları üzerinden, ilgili maddeye ait özellikler, müdahale koşulları, alınması gereken önlemler ve diğer kritik bilgiler veri tabanından alınarak kullanıcılara sunulmuştur. Bu bilgiler, görsel çıktılarla eşleştirilerek uyarı sistemleri aracılığıyla kullanıcıya aktarılmıştır. Çalışmada, tehlikeli madde etiketlerini içeren ve çeşitli arka plan koşullarına sahip 600 adet görselden oluşan bir veri seti kullanılmıştır. Veri seti üzerinde yapılan testlerde, doğruluk metrikleri hesaplanarak sistemin performansı değerlendirilmiştir. Elde edilen sonuçlar, OCR tabanlı yaklaşımın tehlikeli madde etiketlerinin algılanması ve işlenmesindeki etkinliğini göstermiştir. Bu çalışma, özellikle büyük ölçekli lojistik operasyonlarında, güvenli taşıma ve hızlı müdahale süreçleri için önemli bir katkı sunmaktadır.

Keywords

• Görüntü işleme
• tehlikeli madde
• optik karakter tanıma

Dangerous Goods Detection and Warning Approach Based on Image Processing Techniques

Abstract

Hazardous substances are widely used in many sectors such as industry, logistics, agriculture and energy, but they carry potentially serious risks. Accurate identification of these risks before the materials start transportation processes is critical to prevent potential accidents and minimize risks. This study presents an approach to preventing accidents that may occur in the transport of dangerous goods to ensure rapid, effective intervention in case of possible accidents and to take early precautions. Optical Character Recognition (OCR) technology, one of the image processing techniques, is used in the study. Dangerous goods labels were detected with the help of OCR algorithms and the texts on the label were successfully detected. The detected texts, especially the United Nations (UN) numbers specific to hazardous substances, were matched with a previously created database. Based on the UN numbers matched with the database, the properties of the relevant substance, response conditions, precautions to be taken and other critical information were retrieved from the database and presented to the users. This information is matched with visual outputs and transferred to the user through warning systems. In the study, a dataset of 600 images containing hazardous material labels with various background conditions was used. In the tests performed on the dataset, the performance of the system was evaluated by calculating accuracy metrics. The results show the effectiveness of the OCR-based approach in detecting and processing hazardous material labels. This study provides an important contribution for safe transportation and rapid response processes, especially in large-scale logistics operations.

Keywords

• Image Processing
• Dangerous Goods
• Optical Character Recognition

References

1. Ellena LM, Olampi S, Guarnieri F. Technological risks management: Automatic detection and identification of hazardous material transportation trucks. WIT Trans Ecol Environ 2004; 77: 763-771.
2. Yang Q, Chin KS, Li YL. A quality function deployment-based framework for the risk management of hazardous material transportation process. Journal of Loss Prevention in the Process Industries 2018; 52: 81-92.
3. Vojinović N, Sremac S, Zlatanović D. A Novel Integrated Fuzzy‐Rough MCDM Model for Evaluation of Companies for Transport of Dangerous Goods. Complexity 2021; 2021(1): 5141611.
4. Hsu C, Yang J, Chang A, Liu G. A new hybrid MCDM approach for mitigating risks of hazardous material road transportation. Mathematical biosciences and engineering 2024; 21(3): 4210-4240.
5. Liu Y, Fan LS, Li X, Shi SL, Lu Y. Trends of hazardous material accidents (HMAs) during highway transportation from 2013 to 2018 in China. Journal of Loss Prevention in the Process Industries 2020; 66:104150.
6. Panke ZH, Fan LU. Evolutionary game analysis on safety supervision of general aviation based on system dynamics simulation. China Safety Science Journal 2019; 29(4): 43.
7. Parra A, Zhao B, Haddad A, Boutin M, Delp EJ. Hazardous material sign detection and recognition. In: IEEE 2013 International Conference on Image Processing; 15-18 Sep 2013; Melbourne, Australia: IEEE. pp. 2640-2644.
8. Jia L, Wang J, Wang T, Li X, Yu H, Li Q. HMD-net: a vehicle hazmat marker detection benchmark. Entropy 2022; 24(4): 466-481.

9. Pramanik A, Sarkar S, Maiti J. Oil spill detection using image processing technique: An occupational safety perspective of a steel plant. Emerging Technologies in Data Mining and Information Security: Proceedings of IEMIS 2018; 3: 247-257).
10. Monteiro G, Camelo L, Aquino G, Fernandes RD, Gomes R, Printes A, Torné I, Silva H, Oliveira J, Figueiredo C. A Comprehensive Framework for Industrial Sticker Information Recognition Using Advanced OCR and Object Detection Techniques. Applied Sciences 2023; 13(12): 7320-7336.
11. Shahin M, Chen FF, Hosseinzadeh A. Machine-based identification system via optical character recognition. Flexible Services and Manufacturing Journal 2024; 36(2):453-480.
12. Tang Q, Lee Y, Jung H. The Industrial Application of Artificial Intelligence-Based Optical Character Recognition in Modern Manufacturing Innovations. Sustainability 2024; 16(5): 2161-2181.
13. Memon J, Sami M, Khan RA, Uddin M. Handwritten optical character recognition (OCR): A comprehensive systematic literature review (SLR). IEEE access 2020; 8:142642-142668.
14. Okur FB, Eyüpoğlu C. Image matching based hazardous material detection and warning system. Istanbul Ticaret University Fen Bilimleri Dergisi 2024; 23(46): 271-291.
15. Application of RFID Technology in Ship Dangerous Goods Transportation Management, JTSpeedwork. https://www.jtspeedwork.com/application-of-rfid-technology-in-ship-dangerous-goods-transportation management_n101. [Accessed: Jan. 25, 2025].
16. Mishra A, Ram AS. Handwritten Text Recognition Using Convolutional Neural Network. arXiv preprint arXiv:2307.05396 2023.
17. Olson L, Berry V. Digitization decisions: comparing OCR software for librarian and archivist use. Code4Lib Journal 2021; 22(52).
18. Santos Y, Silva M, Reis JC. Evaluation of optical character recognition (OCR) systems dealing with misinformation in Portuguese. In: IEEE 2023 36th SIBGRAPI Conference on Graphics, Patterns and Images (SIBGRAPI); 6-9 Nov 2023; Rio Grande, RS, Brazil: IEEE. pp. 223-228.
19. Palkovič R. Text Recognition in Historic Birth, Death and Marriage Records. Bachelor’s Thesis, Masaryk University, Brno, Czechia, 2023.
20. Hukkeri GS, Goudar RH, Janagond P, Patil PS. Machine Learning in OCR Technology: Performance Analysis of Different OCR Methods for Slide-to-Text Conversion in Lecture Videos. International Journal of Advanced Computer Science and Applications 2022; 13(8): 325-332.
21. Alimovski E, Erdemir G, Kuzucuoglu AE. Text Detection and Recognition in Natural Scenes by Mobile Robot. European Journal of Technique (EJT) 2024; 14(1): 1-7.
22. OCR Accuracy: What Is It and How to Measure It?, Docsumo, Available: https://www.docsumo.com/blogs/ocr/accuracy#:~:text=Good%20OCR%20accuracy%3A%20CER%201,(i.e.%20below%2090%25%20accurate). [Accessed: Jan. 25, 2025].
23. Mohamed MA, Tünnermann J, Mertsching B. Seeing Signs of Danger: Attention-Accelerated Hazmat Label Detection. In: IEEE 2018 International Symposium on Safety, Security, and Rescue Robotics (SSRR); 6-8 Aug 2018; Philadelphia, PA, USA: IEEE. pp. 1-6.
24. Schramm E. Hazmat13 Dataset, Kaggle, 2023. Available: https://www.kaggle.com/datasets/eliasschramm/hazmat13. [Accessed: Jan. 14, 2025]
25. Zhou X, Yao C, Wen H, Wang Y, Zhou S, He W, Liang J. East: an efficient and accurate scene text detector. In: IEEE 2017 Proceedings of the IEEE conference on Computer Vision and Pattern Recognition; 21-26 Jul 2017; Honolulu, HI, USA: IEEE. pp. 5551-5560.
26. Vedhaviyassh DR, Sudhan R, Saranya G, Safa M, Arun D. Comparative analysis of easyocr and tesseractocr for automatic license plate recognition using deep learning algorithm. In: IEEE 2022 6th International Conference on Electronics, Communication and Aerospace Technology; 1-3 Dec 2022: Coimbatore, Tamil Nadu, India: IEEE. pp. 966-971).
27. Salehudin MA, Basah SN, Yazid H, Basaruddin KS, Safar MJ, Som MM, Sidek KA. Analysis of Optical Character Recognition using EasyOCR under Image Degradation. In: IOP Publishing 2023 Journal of Physics: Conference Series; 1 Nov 2023; 2641(1): IOP Publishing. p. 012001.
28. Hamad K, Kaya M. A detailed analysis of optical character recognition technology. International Journal of Applied Mathematics Electronics and Computers 2016; Sep 1(Special Issue-1):244-249.
29. Kissos I, Dershowitz N. OCR error correction using character correction and feature-based word classification. In: IEEE 2016 12th IAPR Workshop on Document Analysis Systems (DAS);11-14 Apr 2016; Santorini, Greece: IEEE. pp. 198-203.
30. Harraj AE, Raissouni N. OCR accuracy improvement on document images through a novel pre-processing approach. Signal & Image Processing: An International Journal 2015; 6(4): 01–18.
31. de Oliveira LL, Vargas DS, Alexandre AM, Cordeiro FC, Gomes DD, Rodrigues MD, Romeu RK, Moreira VP. Evaluating and mitigating the impact of OCR errors on information retrieval. International Journal on Digital Libraries 2023; 24(1): 45-62.
32. Subramani N, Matton A, Greaves M, Lam A. A survey of deep learning approaches for ocr and document understanding. arXiv preprint arXiv:2011.13534 2021; Nov 27.
33. Dölek I, Kurt A. Ottoman Optical Character Recognition with deep neural networks Derin sinir ağlarıyla Osmanlıca optik karakter tanıma. Journal of the Faculty of Engineering and Architecture of Gazi University 2023;38(4): 2579-2593.