Image processing for material research in nuclear energetics
| Thesis title in Czech: | Zpracování obrazu pro materiálový výzkum v jaderné energetice |
|---|---|
| Thesis title in English: | Image processing for material research in nuclear energetics |
| Academic year of topic announcement: | 2019/2020 |
| Thesis type: | dissertation |
| Thesis language: | angličtina |
| Department: | Department of Software and Computer Science Education (32-KSVI) |
| Supervisor: | RNDr. Jan Blažek, Ph.D. |
| Author: | hidden - assigned and confirmed by the Study Dept. |
| Date of registration: | 23.09.2020 |
| Date of assignment: | 23.09.2020 |
| Confirmed by Study dept. on: | 30.09.2020 |
| Advisors: | doc. RNDr. Elena Šikudová, Ph.D. |
| Guidelines |
| The growing availability of digital cameras enlarges their application in industrial control systems, where they fulfill or even replace more expensive sensors (e.g. ultrasound-based or laser-based). The advantage of cameras is especially their data output, images, which are comprehensible for humans. Thanks to that a methodology of data evaluation can be easily created. In the field of material research, they are visible degradations: the most often chemical or geometrical, or various mechanical damages. At this place, a space for image processing algorithms and methodology automation opens. Automation helps with dataset acquisition necessary for the analysis of tested materials or accelerates the classification of object degradation.
The main focus of the work is the selection of suitable algorithms for image processing and development of new ones, which will be appropriate for the detection and classification of visible degradations and material damages and which will enable tracking of their evolution during the time. The student will be able to improve the way of data collection and set up a desirable methodology for quality image-based records acquisition. |
| References |
| Gonzalez, R. C. & Woods, R. E. (2008), Digital image processing , Prentice Hall , Upper Saddle River, N.J. .
Goodfellow et. al. (2016), Deep Learning, MIT Press. Martina Mala. Seven years of inspections on tvsa-t fuel assemblies at temelin npp. 12 International conference on WWER fuel performance, modelling and experimental support, Bulgaria, 2017. Martina Mala and Marek Miklos. Nuclear fuel inspections and repairs at the temelin npp. Bezpecnost Jaderne Energie, 19(11-12):362–365, 2011. Martina Mala. Post-irradiation inspections on tvsa-t fuel assemblies at temel ́ın npp. VVER 2013, 2013. and other relevant scientific papers |
| Preliminary scope of work |
| S rostoucí dostupností digitálních kamer se rozšiřuje i jejich nasazování v průmyslových kontrolních systémech, kde doplňují nebo i nahrazují dražší senzory (například ultrazvukové nebo laserové). Výhodou kamer je zejména jejich datový výstup, obraz, který je snadno srozumitelný pro člověka. Díky tomu poměrně snadno vzniká metodika datového vyhodnocování. V oblasti materiálového výzkumu se jedná o různé viditelné degradace: nejčastěji chemické nebo geometrické, nebo o různá mechanická poškození. Na tomto místě se otevírá prostor pro algoritmy zpracování obrazu a automatizaci celé metodiky, která pomůže při získávání potřebných datasetů pro analýzu testovaných materiálů nebo urychlí klasifikaci degradace objektů.
Náplní práce je výběr vhodných algoritmů pro zpracování obrazu a návrh algoritmů nových, které budou vhodné pro detekci a klasifikaci viditelných degradací a poškození materiálu a umožňující sledování jejich vývoje v čase. Student bude mít možnost ovlivnit i způsob pořizování vstupních dat a sestavit potřebnou metodiku pro získání kvalitních obrazových záznamů. |
| Preliminary scope of work in English |
| The growing availability of digital cameras enlarges their application in industrial control systems, where they fulfill or even replace more expensive sensors (e.g. ultrasound-based or laser-based). The advantage of cameras is especially their data output, images, which are comprehensible for humans. Thanks to that a methodology of data evaluation can be easily created. In the field of material research, they are visible degradations: the most often chemical or geometrical, or various mechanical damages. At this place, a space for image processing algorithms and methodology automation opens. Automation helps with dataset acquisition necessary for the analysis of tested materials or accelerates the classification of object degradation.
The main focus of the work is the selection of suitable algorithms for image processing and development of new ones, which will be appropriate for the detection and classification of visible degradations and material damages and which will enable tracking of their evolution during the time. The student will be able to improve the way of data collection and set up a desirable methodology for quality image-based records acquisition. |