Development of an AI system for automated detection and classification of hazards in x-ray facilities

Services

Business Consulting

Custom Software

Industries

Custom Software

Artificial Intelligence

Client

a technical consultant company

Period

April 2020 - March 2022

The aim of the project was to develop an advanced AI system for the automated detection and classification of hazards in fluoroscopy systems. A key component of our work is the implementation of pre-processing of the image data to provide an optimal basis for the AI analysis. This included cleaning, normalising and, where necessary, transforming the raw data to reduce noise and improve the accuracy of the subsequent AI models. We also designed and implemented custom AI models based on the specific requirements and conditions of the screening facilities. These models were trained to detect and classify potential hazards with high accuracy and reliability. In parallel, we worked on the design and construction of embedded hardware architectures to ensure efficient integration and execution of the AI algorithms in the systems.

Credits

Lukas Famula

Technologies

C#
Python
NVIDIA SDK
PyTorch
TensorFLow
Linux Server
Docker
Git