Deep 3D Imaging

As individuals, our ability to acquire digital images of objects in the world around us has grown dramatically in recent years.

As individuals, our ability to acquire digital images of objects in the world around us has grown dramatically in recent years. Using powerful computational tools, we can easily merge the information from many digital pictures into 3D models. In comparison, the field of 3D interior scanning (using CT, MRI, ultrasound, and so on) is still dominated by big machines that can only be operated by expert users using strict protocols. Our challenge is to develop mathematical techniques and tools that make it possible to acquire deep 3D images in a far more flexible way, using new, compact scanner designs that can image quickly, safely, and at high resolutions.

By using a broad range of mathematical techniques from analysis and scientific computing, we create algorithms that can compute accurate images of the interior of objects from highly limited measurement data. There is a real need for such solutions in science (for improved imaging of materials and biological specimens), industry (for more flexible deep interior quality inspection), and medicine (reducing harmful radiation dose in scans).

The Computational Imaging group develops fundamentally new algorithms for 3D-image reconstruction and turns these into open software that is used by academic users in the application fields, as well as by companies for integration into commercial products.

Contact person: Joost Batenburg
Research group: Computational Imaging (CI)
Research partners: Rijksmuseum, Naturalis, FEI Company