Quantitative image analysis: a focus on automated characterization of structures in optical microscopy of iron ore pellets
2013 (English)Licentiate thesis, comprehensive summary (Other academic)
Sintering occurs in many types of material such as iron, ceramics and snow, typically during thermal treatment, and aects the material properties, particularly the strength, by the bonding of particles into a coherent structure. In order to improve the mechanical strength in magnetite iron ore pellets it is important to be able to characterize and quantitatively measure the degree of sintering and features that impact the process of sintering.The aim for this licentiate thesis has been to create tools for sintering characterization through automated image analysis of optical microscopy images. Such tools are of interest since they provide a comparable quantication of pellet properties that can be related to other parameters, giving a historical record that is digital, objective and not dependent on the eyes of a trained expert. In this work, two dierent studies of the microstructure in indurated (heat hardened) pellets have been performed. The methods presented in these studies have been shown suitable for characterizing sintering properties in iron ore pellets, and possibly also other materials that experience sintering phenomena.The first study presents research to automate image capture and analysis of entire crosssections of indurated iron ore pellets to characterize proportions of magnetite, hematite, and other components. Spatial distributions of the mentioned phases are produced for each pellet, graphing proportions in relation to the distance to the pellet surface. The results are not directly comparable to a chemical analysis but comparisons with manual segmentation of images validates the method. Dierent types of pellets have been tested and the system has produced robust results for varying cases.The second study focuses on the analysis of the particle joins and structure. The joins between particles have been identied with a method based mainly on morphological image processing and features have been calculated based on the geometric properties and curvature of these joins. The features have been analyzed and been determined to hold discriminative power by displaying properties consistent with sintering theory and results from traditional physical dilation measurements on the heated samples.A note of caution for quantitative studies of iron ore pellet has been identied in this thesis. Especially for green pellets, the microscopy sample preparation prohibit any statistical inference studies due to particle rip-out during polishing. Researchers performing qualitative microscopy studies are generally aware of the phenomenon of rip-outs, but the extent of how even seemingly good samples are aected has not been unveiled until attempting extensive quantitative analysis of features such as green pellet porosity during the course of this work.
Place, publisher, year, edition, pages
Luleå: Luleå tekniska universitet, 2013.
Licentiate thesis / Luleå University of Technology, ISSN 1402-1757
Research subject Signal Processing
IdentifiersURN: urn:nbn:se:ltu:diva-26204Local ID: d2915986-de44-485b-aebf-eb6e6e1c98ccISBN: 978-91-7439-585-3ISBN: 978-91-7439-586-0 (PDF)OAI: oai:DiVA.org:ltu-26204DiVA: diva2:999364
Godkänd; 2013; 20130224 (frinel); Tillkännagivande licentiatseminarium 2013-04-25 Nedanstående person kommer att hålla licentiatseminarium för avläggande av teknologie licentiatexamen. Namn: Frida Nellros Ämne: Signalbehandling/Signal Processing Uppsats: Quantitative Image Analysis – A Focus on Automated Characterization of Structures in Optical Microscopy of Iron Ore Pellets Examinator: Associate Professor Matthew Thurley, Institutionen för system- och rymdteknik, Luleå tekniska universitet Diskutant: Associate Professor Carolina Wählby, Centrum for Image Analysis, Uppsala universitet Tid: Fredag den 17 maj 2013 kl 12.30 Plats: A109, Luleå tekniska universitet2016-09-302016-09-30Bibliographically approved