Automated X-ray image analysis for cargo security: Critical review and future promise

Rogers, Thomas W.; Jaccard, Nicolas; Morton, Edward J.; Griffin, Lewis D.

doi:10.3233/xst-160606

Cited by 47 publications

(33 citation statements)

References 64 publications

(156 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Computer Aided Screening (CAS) performs automated threat detection in the generalized sense, however this largely remains an unsolved problem. Previous work [21], [22] has focused on image enhancement [23]- [25], segmentation [26], [27], classification [1], [2], [4], [6], [17] or detection [5], [10], [28], [29] tasks in order to further investigate the real time applicability of CAS to automatize aviation security screening. For a detailed overview the reader is directed to Rodgers et al [22] and Mouton et al [21].…”

Section: Related Workmentioning

confidence: 99%

Using Deep Convolutional Neural Network Architectures for Object Classification and Detection Within X-Ray Baggage Security Imagery

Akçay

Kundegorski

Willcocks

et al. 2018

IEEE Trans.Inform.Forensic Secur.

297

180

View full text Add to dashboard Cite

We consider the use of deep Convolutional Neural Networks (CNN) with transfer learning for the image classification and detection problems posed within the context of X-ray baggage security imagery. The use of the CNN approach requires large amounts of data to facilitate a complex end-to-end feature extraction and classification process. Within the context of Xray security screening, limited availability of object of interest data examples can thus pose a problem. To overcome this issue, we employ a transfer learning paradigm such that a pre-trained CNN, primarily trained for generalized image classification tasks where sufficient training data exists, can be optimized explicitly as a later secondary process towards this application domain. To provide a consistent feature-space comparison between this approach and traditional feature space representations, we also train Support Vector Machine (SVM) classifier on CNN features. We empirically show that fine-tuned CNN features yield superior performance to conventional hand-crafted features on object classification tasks within this context. Overall we achieve 0.994 accuracy based on AlexNet features trained with Support Vector Machine (SVM) classifier. In addition to classification, we also explore the applicability of multiple CNN driven detection paradigms such as sliding window based CNN (SW-CNN), Faster RCNN (F-RCNN), Region-based Fully Convolutional Networks (R-FCN) and YOLOv2. We train numerous networks tackling both single and multiple detections over SW-CNN/F-RCNN/R-FCN/YOLOv2 variants. YOLOv2, Faster-RCNN, and R-FCN provide superior results to the more traditional SW-CNN approaches. With the use of YOLOv2, using input images of size 544×544, we achieve 0.885 mean average precision (mAP) for a six-class object detection problem. The same approach with an input of size 416×416 yields 0.974 mAP for the two-class firearm detection problem and requires approximately 100ms per image. Overall we illustrate the comparative performance of these techniques and show that object localization strategies cope well with cluttered X-ray security imagery where classification techniques fail.

show abstract

Section: Related Workmentioning

confidence: 99%

Using Deep Convolutional Neural Network Architectures for Object Classification and Detection Within X-Ray Baggage Security Imagery

Akçay

Kundegorski

Willcocks

et al. 2018

IEEE Trans.Inform.Forensic Secur.

297

180

View full text Add to dashboard Cite

show abstract

“…The urgent need for robust methods to fill the detection capability gap is not being matched by the current research output in automated analysis for X-ray cargo images, which was recently throughly documented and reviewed in Ref. [13]. Impressive performance has been reported for the detection of security threats (including SMTs) [7,[14][15][16] in baggage X-ray images, partly made possible by the small dimensions and complexity (e.g.…”

Section: Related Workmentioning

confidence: 99%

Automated detection of smuggled high-risk security threats using Deep Learning

Jaccard

Rogers

Morton³

et al. 2016

7th International Conference on Imaging for Crime Detection and Prevention (ICDP 2016)

View full text Add to dashboard Cite

The security infrastructure is ill-equipped to detect and deter the smuggling of non-explosive devices that enable terror attacks such as those recently perpetrated in western Europe. The detection of so-called "small metallic threats" (SMTs) in cargo containers currently relies on statistical risk analysis, intelligence reports, and visual inspection of X-ray images by security officers. The latter is very slow and unreliable due to the difficulty of the task: objects potentially spanning less than 50 pixels have to be detected in images containing more than 2 million pixels against very complex and cluttered backgrounds. In this contribution, we demonstrate for the first time the use of Convolutional Neural Networks (CNNs), a type of Deep Learning, to automate the detection of SMTs in fullsize X-ray images of cargo containers. Novel approaches for dataset augmentation allowed to train CNNs from-scratch despite the scarcity of data available. We report fewer than 6% false alarms when detecting 90% SMTs synthetically concealed in streamof-commerce images, which corresponds to an improvement of over an order of magnitude over conventional approaches such as Bag-of-Words (BoWs). The proposed scheme offers potentially super-human performance for a fraction of the time it would take for a security officers to carry out visual inspection (processing time is approximately 3.5s per container image).

show abstract

“…Nowadays, baggage inspection is mostly done manually by looking at the acquired images, but there is an increasing interest in automated methods for detecting threat items. 4 Until now, there has been limited academic interest in developing automated methods for inspection, which might be due to limited access to data 6 and a lack of regulatory requirements. Data is essential for developing methods for detection of objects like handguns, knives, etc.…”

Section: Introductionmentioning

confidence: 99%

Multispectral x-ray CT: multivariate statistical analysis for efficient reconstruction

Kheirabadi

Mustafa

Lyksborg

et al. 2017

Developments in X-Ray Tomography XI

View full text Add to dashboard Cite

Recent developments in multispectral X-ray detectors allow for an efficient identification of materials based on their chemical composition. This has a range of applications including security inspection, which is our motivation. In this paper, we analyze data from a tomographic setup employing the MultiX detector, that records projection data in 128 energy bins covering the range from 20 to 160 keV. Obtaining all information from this data requires reconstructing 128 tomograms, which is computationally expensive. Instead, we propose to reduce the dimensionality of projection data prior to reconstruction and reconstruct from the reduced data. We analyze three linear methods for dimensionality reduction using a dataset with 37 equally-spaced projection angles. Four bottles with different materials are recorded for which we are able to obtain similar discrimination of their content using a very reduced subset of tomograms compared to the 128 tomograms that would otherwise be needed without dimensionality reduction.

show abstract

Automated X-ray image analysis for cargo security: Critical review and future promise

Cited by 47 publications

References 64 publications

Using Deep Convolutional Neural Network Architectures for Object Classification and Detection Within X-Ray Baggage Security Imagery

Using Deep Convolutional Neural Network Architectures for Object Classification and Detection Within X-Ray Baggage Security Imagery

Automated detection of smuggled high-risk security threats using Deep Learning

Multispectral x-ray CT: multivariate statistical analysis for efficient reconstruction

Contact Info

Product

Resources

About