Direkt zum Inhalt springen
login.png Login join.png Register    |
de | en
MyTUM-Portal
Technical University of Munich

Technical University of Munich

Sitemap > Bulletin Board > Diplomarbeiten, Bachelor- und Masterarbeiten > Abschlussarbeiten, Bachelor- und Masterarbeiten-Automatic microscopic image/video analysis using machine learning
up   Back to  News Board    previous   Browse in News  next    

Abschlussarbeiten, Bachelor- und Masterarbeiten-Automatic microscopic image/video analysis using machine learning

19.10.2021, Diplomarbeiten, Bachelor- und Masterarbeiten

Microscopy is one of the most important techniques in many fields. For example, in materials science and engineering, researchers and engineers have for centuries relied on different types of microscopies to closely characterize materials microstructure, surface morphology, and internal defects. This detailed knowledge of materials structure is essential for establishing the so-called materials structure-processing-property relationships, which are the key to materials design and optimization.

At the chair of Materials Engineering of Additive Manufacturing (AM), we concentrate on fast-developing novel metallic materials, especially for AM. Our target is first time right material and AM process solutions for specific applications. We are equipped with light microscopes and a high-temperature laser-scanning confocal microscopy, which produces large microscopic image/video datasets that allow us to track materials solidification and phase transformation in-situ. Usually, scientists analyze these large microscopic data manually, frame-by-frame, which is tedious and time-consuming, becoming a bottleneck for fast materials development. Therefore, the goal of this master thesis is to develop automatic methods to analyze microscopic images and videos using machine learning algorithms. You need to use image processing libraries to preprocess micrographs and build machine learning pipelines using technologies such as neural network to extract microstructural features, e.g., secondary precipitates. Ultimately a quantitative description of their geometrical and morphological characteristics in a statistical manner is expected.

Your profile:
• Programming skills in Python, Matlab, or related languages
• Knowledge of materials microstructure and phase transformation
• Knowledge of microscopy and microanalysis
• Knowledge of machine learning and imaging processing would be a plus
• Familiar with Numpy, Pandas, scikit-learn, PyTorch, etc. would be a plus

Our offer:
• Necessary equipment
• Positive work environment
• Flexible working hours

Contact:
Dr.-Ing. Zirong Peng
zirong.peng@tum.de
+49 89 289 55346

Kontakt: zirong.peng@tum.de

Todays events

no events today.

Calendar of events