Optical Image Sensing and Analysis of Iron Ore Pellets: A Machine Learning Approach

Year : 2025 | Volume : 15 | Issue : 03 | Page : 7 18
    By

    Rakesh Prasad,

  • Lokesh Pachauri,

  • Alok Singh,

  • Madhav Rawat,

  • Hardeep Singh,

  • Vijay S. Katta,

  1. Assistant Professor, Mechanical Engineering Department, Hindustan College of Science & Technology, Mathura, Uttar Pradesh, India
  2. Student, Computer Science & Engineering Department, Hindustan College of Science & Technology, Mathura, Uttar Pradesh, India
  3. Student, Computer Science & Engineering Department, Hindustan College of Science & Technology, Mathura, Uttar Pradesh, India
  4. Student, Computer Science & Engineering Department, Hindustan College of Science & Technology, Mathura, Uttar Pradesh, India
  5. Student, Computer Science & Engineering Department, Hindustan College of Science & Technology, Mathura, Uttar Pradesh, India
  6. Assistant Professor, Computer Science & Engineering Department, Hindustan College of Science & Technology, Mathura, Uttar Pradesh, India

Abstract

The present work is aimed to improve quality control in steel production using SEM imaging and machine learning. High-resolution SEM images of iron ore pellets, primarily composed of hematite and magnetite, are analyzed to understand their microstructural features, which significantly impact pellet performance during reduction processes. Traditional microstructure analysis is manual, time- consuming, and prone to inconsistencies. This study proposes an automated approach using K-Means Clustering, Canny Edge Detection, DBSCAN, and manual pixel-based testing to classify microstructures effectively. K-Means segments SEM images based on pixel intensity, identifying different material phases. Canny Edge Detection is used to highlight sharp boundaries and fine structures in the images, improving the clarity of segmentation. DBSCAN further enhances classification by detecting dense microstructural regions and filtering out noise and outliers. Together, these techniques automate and improve the accuracy of microstructure analysis. This integrated method provides reliable insights into the relationship between microstructure and pellet quality. The project ultimately aims to optimize pellet production, reduce inspection time, lower operational costs, and enhance the efficiency of steel manufacturing processes.

Keywords: Optical sensing, energy band, microstructure, image processing, K-mean, DBSCAN

[This article belongs to Journal of Instrumentation Technology & Innovations ]

How to cite this article:
Rakesh Prasad, Lokesh Pachauri, Alok Singh, Madhav Rawat, Hardeep Singh, Vijay S. Katta. Optical Image Sensing and Analysis of Iron Ore Pellets: A Machine Learning Approach. Journal of Instrumentation Technology & Innovations. 2025; 15(03):7-18.
How to cite this URL:
Rakesh Prasad, Lokesh Pachauri, Alok Singh, Madhav Rawat, Hardeep Singh, Vijay S. Katta. Optical Image Sensing and Analysis of Iron Ore Pellets: A Machine Learning Approach. Journal of Instrumentation Technology & Innovations. 2025; 15(03):7-18. Available from: https://journals.stmjournals.com/joiti/article=2025/view=227622


References

  1.  Sarkar A, Chavan V, Pai NN, Prakash A, Hazra B, Raut P, Sunilkumar D, Sivananda C, Kundu S, Nag S, Basu S. Reduction of Iron Ore Pellets: A Microstructural Perspective? Metall Mater Trans A. 2024 Feb; 55(2): 537–49.
  2. Meyer M, Lagoeiro LE, Graça LM, Silva CJ. Phase and microstructural characterization of iron ore pellet and their relation with cold crushing strength test. Miner Process Extr Metall Rev. 2016 Sep 2; 37(5): 295–304.
  3. Prasad R, Gupta K, Poras A, Agarwal S. Microstructure and phase analysis of indurated iron ore pellets. Mater Today: Proc. 2020 Jan 1; 26: 2636–42.
  4. Prasad R, Venugopal R, Kumaraswamidhas LA, Pandey C, Pan SK. Analysis of the influence of Blaine numbers and firing temperature on iron ore pellets properties using RSM-I-Optimal design: an approach toward suitability. Min Metall Explor. 2020 Oct; 37: 1703–16.
  5. Prasad R, Soren S, Kumaraswamidhas LA, Pandey C, Pan SK. Experimental investigation of different fineness and firing temperatures on pellets properties of different iron ore fines from Indian mines. Materials. 2022 Jun 14; 15(12): 4220.
  6. Sharma T, Gupta RC, Prakash B. Effect of reduction rate on the swelling behaviour of iron ore pellets. ISIJ Int. 1992;32(7):812-818. doi:10.2355/isijinternational.32.812.
  7. Umadevi T, Kumar P, Lobo NF, Mahapatra PC, Prabhu M, Ranjan M. Effect of iron ore pellet size on its properties and microstructure. Steel Res Int. 2009 Oct; 80(10): 709–16.
  8. Chowdhury A, Kautz E, Yener B, Lewis D. Image driven machine learning methods for microstructure recognition. Comput Mater Sci. 2016 Oct 1; 123: 176–87.
  9. Campbell A, Murray P, Yakushina E, Marshall S, Ion W. New methods for automatic quantification of microstructural features using digital image processing. Mater Des. 2018 Mar 5; 141: 395–406.
  10. Suyala M, Sharmab S. A Review on Analysis of K-Means Clustering Machine Learning Algorithm based on Unsupervised Learning. J Artif Intell Syst. 2024; 6(1): 85–95.
  11. hen , iu X, a J, Zhao , ou X arameter selection algorithm o based on ‐ means two classification algorithm. J Eng. 2019 Dec; 2019(23): 8676–9.
  12. Wu F, Zhu C, Xu J, Bhatt MW, Sharma A. Research on image text recognition based on canny edge detection algorithm and k-means algorithm. Int J Syst Assur Eng Manag. 2022 Mar; 13(Suppl 1): 72–80.
  13. Aversa R, Coronica P, De Nobili C, Cozzini S. Deep learning, feature learning, and clustering analysis for SEM image classification. Data Intell. 2020 Oct 1; 2(4): 513–28.
  14. Bezerra ET, Augusto KS, Paciornik S. Discrimination of pores and cracks in iron ore pellets using deep learning neural networks. REM-Int Eng J. 2020 Apr 17; 73: 197–203.
  15. Deo AJ, Sahoo A, Behera SK, Das DP. Iron ore pellet size analysis: A machine learning-based image processing approach. IEEE Ind Appl Mag. 2022 Nov 24; 29(1): 67–79.
  16. Tang , a ang Y, iu Y, one ands T , ’ ea , ostaghimi , rmstrong T, nac stedt M. Particle classification of iron ore sinter green bed mixtures by 3D X-ray microcomputed tomography and machine learning. Powder Technol. 2023 Feb 1; 415: 118151.
  17. Mancini M, Mircoli A, Potena D, Diamantini C, Duca D, Toscano G. Prediction of pellet quality through machine learning techniques and near-infrared spectroscopy. Comput Ind Eng. 2020 Sep 1; 147: 106566.
Regular Issue Subscription Original Research
Volume 15
Issue 03
Received 23/05/2025
Accepted 26/05/2025
Published 26/07/2025
Publication Time 64 Days
193

Login


My IP

PlumX Metrics