Afbeelding
CVDS-PROJECT-XRAY

Contraband anomaly detection

As part of the Professorship of Applied Sciences in Computer Vision and Data Science

Goal

Together with the company OD Security detect contraband in x-ray images.

Main challenge  

Contraband shape and features are not known in advance and there is
limited data available. To limit the hazardous x-ray exposure the data
has a low signal-to-noise ratio. 

Result 

An anomaly detection AI model to detect previously unseen contraband

Future

Further improvement of the anomaly detection algorithms and investigate
this performance in other applications.