Intelligent train inspection
Image credit:
This project was undertaken during my HIWI position at Fraunhofer and was designed for presentation at the Innotrans exhibition. The objective was to develop an innovative system capable of scanning trains using an industrial camera, processing the acquired data, and detecting alterations, specifically graffiti, on the train surfaces. The project involved a comprehensive hardware-software interaction utilizing a Raspberry Pi as the central processing unit. I implemented a graphical user interface (GUI) using Flask, enabling users to interact with the system through a web-based platform.
One of the key challenges of this project was the highly specialized nature of the use case, which presented unique problems without readily available solutions. To address these challenges, I employed creative problem-solving techniques, designing tailored algorithms for image processing and analysis to ensure accurate detection of graffiti and other changes. This project not only highlights the intersection of computer vision and real-time data processing but also emphasizes the importance of innovation in developing effective solutions for specific industrial applications.
