National
Security
X-ray
Baggage Imagery Analysis
Airport
Carry-on Baggage Checkpoints
With increased security needs for vigilance, efficiency, and accuracy
in detecting potential threat items in carry-on baggage, there has been
a dramatic increase in the quantity of x-ray systems deployed in airports
nationwide. Coupled with this is a renewed national interest in the exploration
of new technologies for operator training and assistance. Researchers
at ORNL have been exploring computer vision techniques for assisting
x-ray screeners in identifying potential threat items in carry-on baggage.
Base
Technology
Methods
for creating a searchable image database for queries based on
image content that were developed for the semiconductor industry
were modified and adapted to the problem of identifying and recognizing
parts of threat items found in x-ray baggage imagery. These threats
can include conventional weapons such as guns or knives as well
as improvised explosive devices. The ORNL technology is built
around a database of image characteristics of regions identified
to be threats along with regions known to contain common, harmless
carry-on items. By comparing a suspect area in an image to a
database of known and labeled regions, an improved analysis of
the threat-potential for this item can be performed.
Specifications
and Features
- Method
for operator assistance in identifying potential conventional
threats in x-ray imagery
- Use
of imagery characteristics in dual energy images to confirm
or deny that a suspicious organic material is an explosive
- Windows
software aide to familiarize screeners with the appearance
of threat items in a variety of poses and in the presence of
clutter
Fact
Sheet available here in PDF format.
Point
of Contact:
Regina Ferrell, Development Staff
Image Science & Machine Vision
Engineering Science and Technology Division
Oak Ridge National Laboratory
P.O. Box 2008, MS-6121
Oak Ridge, Tennessee 37831-6121
Office: (865) 574-5730
E-mail: ferrellrk@ornl.gov |
