The proof of concept of a fused radiometric and optical stereoscopic imaging system
The proof of concept of a fused radiometric and optical stereoscopic imaging device is presented. The project was in collaboration with the National Nuclear Laboratory and the Nuclear Decommissioning Authority with the aim of developing a sensor that can be deployed in a nuclear decommissioning envi...
| Main Author: | |
|---|---|
| Published: |
University of Liverpool
2014
|
| Subjects: | |
| Online Access: | http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.639439 |
| Summary: | The proof of concept of a fused radiometric and optical stereoscopic imaging device is presented. The project was in collaboration with the National Nuclear Laboratory and the Nuclear Decommissioning Authority with the aim of developing a sensor that can be deployed in a nuclear decommissioning environment. The radiometric system was a Compton camera comprised of two HPGe planar detectors and presents a significant improvement in efficiency and dynamic range over coded aperture systems currently used in industry. The optical stereoscopic camera is the proprietary Bumblebee XB3 system that provides 3D physical information of the surroundings. Two main experiments are presented; the first investigated the disparity between true source location and reconstructed image position. This disparity was proven and methods for accounting for and correcting it were developed, whereby the image position accuracy was improved by a factor of 26.7. The second experiment imaged 20 MBq $^{137}$Cs sources at distances of 80 - 150 cm with both radiometric and optical stereoscopic systems simultaneously. The first fused images were produced using this data, with the radiometric sources and surroundings clearly visible. A GUI was developed in Matlab to process and fuse the data. Alongside both experiments image optimisation techniques were investigated. Pulse shape analysis was implemented and shown to improve image resolution by 30\% on average at the expense of efficiency. Fold 2 event imaging was conversely shown to improve efficiency at the expense of image resolution. This work provides the basis to develop the project towards a complete system. The steps that must be taken to realise this are outlined and recommendations for overcoming potential challenges are discussed. |
|---|