eduzhai > Applied Sciences > Engineering >

Multi-Class 3D Object Detection Within Volumetric 3D Computed Tomography Baggage Security Screening Imagery

  • king
  • (0) Download
  • 20210507
  • Save

... pages left unread,continue reading

Document pages: 8 pages

Abstract: Automatic detection of prohibited objects within passenger baggage isimportant for aviation security. X-ray Computed Tomography (CT) based 3Dimaging is widely used in airports for aviation security screening whilst priorwork on automatic prohibited item detection focus primarily on 2D X-rayimagery. These works have proven the possibility of extending deepconvolutional neural networks (CNN) based automatic prohibited item detectionfrom 2D X-ray imagery to volumetric 3D CT baggage security screening imagery.However, previous work on 3D object detection in baggage security screeningimagery focused on the detection of one specific type of objects (e.g., either{ it bottles} or { it handguns}). As a result, multiple models are needed ifmore than one type of prohibited item is required to be detected in practice.In this paper, we consider the detection of multiple object categories ofinterest using one unified framework. To this end, we formulate a morechallenging multi-class 3D object detection problem within 3D CT imagery andpropose a viable solution (3D RetinaNet) to tackle this problem. To enhance theperformance of detection we investigate a variety of strategies including dataaugmentation and varying backbone networks. Experimentation carried out toprovide both quantitative and qualitative evaluations of the proposed approachto multi-class 3D object detection within 3D CT baggage security screeningimagery. Experimental results demonstrate the combination of the 3D RetinaNetand a series of favorable strategies can achieve a mean Average Precision (mAP)of 65.3 over five object classes (i.e. { it bottles, handguns, binoculars,glock frames, iPods}). The overall performance is affected by the poorperformance on { it glock frames} and { it iPods} due to the lack of data andtheir resemblance with the baggage clutter.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×