Parametric testing of explosives detection systems using synthetic data augmentation

Ronald Krauss

doi:10.1117/12.2585918

12 April 2021 Parametric testing of explosives detection systems using synthetic data augmentation

Ronald Krauss

Proceedings Volume 11738, Anomaly Detection and Imaging with X-Rays (ADIX) VI; 117380L (2021) https://doi.org/10.1117/12.2585918
Event: SPIE Defense + Commercial Sensing, 2021, Online Only

Abstract

This paper describes a method to acquire actual image data of any material, which could be an explosive or parametric simulant, extract the X-ray features of the object, modify the images in a variety of controlled ways, and record new images. All of the augmented images can then be run through the emulator of the baggage screening system to explore the detection algorithm’s performance. This is similar to the method that was demonstrated over a decade ago, where the Transportation Security Laboratory developed a series of explosive simulants which spanned the X-ray feature space of mass density and effective atomic number, thus demonstrating the capability to effectively map out the performance of explosives detection systems in that feature space. In this paper, the concept has been expanded to include synthetic data, which is far more efficient than creating fully synthetic data and greatly diminishes the need to collect data using real explosives.

Conference Presentation

Citation Download Citation

Ronald Krauss "Parametric testing of explosives detection systems using synthetic data augmentation", Proc. SPIE 11738, Anomaly Detection and Imaging with X-Rays (ADIX) VI, 117380L (12 April 2021); https://doi.org/10.1117/12.2585918

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available