Options
2016
Conference Paper
Titel
Integration of CT methods into hand luggage scanning processes at airport checkpoints based on existing multi-view scanning systems
Abstract
In this presentation, an efficient way to integrate Computed Tomography (CT) methods into the scanning process of carry-on baggage at airport checkpoints is proposed. The implementation suggested is designed to employ existing multi-view scanner geometries. The first key element of this approach addresses the problem of the optimal scanning geometry and focuses on the apt choice of positioning as few as possible additional projection views along the conveyor belt. The second key element is the choice of a suitable and efficient algebraic reconstruction algorithm and its optimization for the purpose of luggage scanning. The main challenge addressed in this investigation is to enhance the quality of the reconstruction of strongly underdetermined data sets with scanned objects whose composition is mostly unknown. This is clearly the case for hand luggage and its very wide ranged possibilities of content. Further technical challenges like the minimization of the reconstruction runtime to in-line capability and several methods to reduce typical artifacts are discussed. Artifact reduction methods shown in the presentation are a, for the used algorithm specialized, metal artifacts reduction (MAR). Also the reduction of noise influence or other image impairments is faced with a splatting type regularization method. An approach to streak artifacts reduction via an individual correction term weighting method is presented. Finally a practical application of the reconstruction results is given where the reconstructed volume is employed to enhance the typical 2D images operators are used to interpret. This is accomplished by removing certain parts of the volume, providing a step towards the ambition that objects like laptops, electronics or other heavy absorbers can be left in the baggage.
Author(s)