Title :
Fully automatic 3D Threat Image Projection: Application to densely cluttered 3D Computed Tomography baggage images
Author :
Megherbi, Najla ; Breckon, Toby P. ; Flitton, Greg T. ; Mouton, Andre
Author_Institution :
Sch. of Eng., Cranfield Univ., Cranfield, UK
Abstract :
In this paper, we describe a Threat Image Projection (TIP) method designed for 3D Computed Tomography (CT) screening systems. The novel methodology automatically determines a valid 3D location in the passenger 3D CT baggage image into which a fictional threat 3D image can be inserted without violating the bag content. According to the scan orientation, the passenger bag content and the material of the inserted threat appropriate CT artefacts are generated using a Radon transform in order to make the insertion realistic. Densely cluttered 3D CT baggage images are used to validate our method. Experimental results confirm that our method is able to reliably insert threat items in challenging 3D images without providing any perceptible visual cue to human screeners.
Keywords :
Radon transforms; aerospace computing; aerospace safety; image processing; inspection; 3D threat image projection; CT screening system; Radon transform; TIP method; aviation security; densely cluttered 3D computed tomography baggage image; passenger bag content; scan orientation; Computed tomography; Image reconstruction; Image segmentation; Materials; Metals; Noise; X-ray imaging; Aviation Security; Computed Tomography; Image processing applications; Threat Image Projection;
Conference_Titel :
Image Processing Theory, Tools and Applications (IPTA), 2012 3rd International Conference on
Conference_Location :
Istanbul
Print_ISBN :
978-1-4673-2585-1
DOI :
10.1109/IPTA.2012.6469523