An evaluation of region based object detection strategies within X-ray baggage security imagery
Title | An evaluation of region based object detection strategies within X-ray baggage security imagery |
Publication Type | Conference Paper |
Year of Publication | 2017 |
Authors | Akcay, S., Breckon, T. P. |
Conference Name | 2017 IEEE International Conference on Image Processing (ICIP) |
Keywords | class firearm detection problem, cluttered X-ray security imagery, Collaboration, convolution, cyber physical systems, Deep Learning, Faster Region-based CNN, feature extraction, first-stage region proposal, Fully Convolutional Networks, image classification, ImageNet dataset, learning (artificial intelligence), Metrics, Microsoft Windows, neural nets, neural networks security, object class X-ray detection problem, object detection, object localization strategies, policy, policy-based governance, Policy-Governed Secure Collaboration, Proposals, pubcrawl, R-CNN/R-FCN variants, region based object detection, resilience, Resiliency, ResNet-101, security, sliding window driven CNN approach, SWCNN, Task Analysis, transfer learning, VGG16, Weapons, X-ray baggage security, X-ray baggage security imagery, X-ray imaging |
Abstract | Here we explore the applicability of traditional sliding window based convolutional neural network (CNN) detection pipeline and region based object detection techniques such as Faster Region-based CNN (R-CNN) and Region-based Fully Convolutional Networks (R-FCN) on the problem of object detection in X-ray security imagery. Within this context, with limited dataset availability, we employ a transfer learning paradigm for network training tackling both single and multiple object detection problems over a number of R-CNN/R-FCN variants. The use of first-stage region proposal within the Faster RCNN and R-FCN provide superior results than traditional sliding window driven CNN (SWCNN) approach. With the use of Faster RCNN with VGG16, pretrained on the ImageNet dataset, we achieve 88.3 mAP for a six object class X-ray detection problem. The use of R-FCN with ResNet-101, yields 96.3 mAP for the two class firearm detection problem requiring 0.1 second computation per image. Overall we illustrate the comparative performance of these techniques as object localization strategies within cluttered X-ray security imagery. |
URL | https://ieeexplore.ieee.org/document/8296499/ |
DOI | 10.1109/ICIP.2017.8296499 |
Citation Key | akcay_evaluation_2017 |
- security
- policy-based governance
- Policy-Governed Secure Collaboration
- Proposals
- pubcrawl
- R-CNN/R-FCN variants
- region based object detection
- resilience
- Resiliency
- ResNet-101
- Policy
- sliding window driven CNN approach
- SWCNN
- Task Analysis
- transfer learning
- VGG16
- Weapons
- X-ray baggage security
- X-ray baggage security imagery
- X-ray imaging
- image classification
- cluttered X-ray security imagery
- collaboration
- convolution
- cyber physical systems
- deep learning
- Faster Region-based CNN
- feature extraction
- first-stage region proposal
- Fully Convolutional Networks
- class firearm detection problem
- ImageNet dataset
- learning (artificial intelligence)
- Metrics
- microsoft windows
- neural nets
- neural networks security
- object class X-ray detection problem
- object detection
- object localization strategies