Under-vehicle imaging systems (UVIS) have proven to be critical for checkpoints that require high security. Yet despite this not all UVIS are created equally. We take a look at how Rapiscan | AS&E’s Gatekeeper Intelligent Vehicle Undercarriage Scanner (IVUS) performs.
Unfortunately, many under-vehicle screening systems on the market do not add value beyond the antiquated inspection technique of viewing the underside of a vehicle with a mirror on a stick. Yet, when designed correctly, an imaging system can scan the underside of a vehicle and produce consistent nighttime and daytime imagery, allowing for foreign object detection to help to quickly identify threats with much higher accuracy and in less time than other inspection methods.
Change Detection is what matters
Manual human inspection of vehicle undercarriages is often ineffective and inefficient. Limitations on humans’ ability to focus on difficult tasks for long durations of time can significantly reduce the likelihood that a security officer will successfully identify a security threat under a vehicle. Beyond the limitations of a person’s attention span, vehicle undercarriages vary significantly among models, making it difficult for a security official to locate threats among the large variety of components and variations. A UVIS uses change detection technology to compare alterations to the undercarriage of a vehicle to reliably identify potential threats. A UVIS should produce high resolution, consistent imagery in all lighting conditions, and the images should be easy to interpret with manual visual inspection on a computer screen.
Simultaneously, change detection will identify modifications to the undercarriage (such as an enlarged fuel tank) by circling or highlighting the area that requires further visual inspection of the image. Since a foreign object can be any form, shape or size, the most effective method of identifying an undefined foreign object is by detecting changes in the imagery (change detection) from a reference database. The Gatekeeper IVUS performs change detection by comparing the underside of the vehicle with a reference image of the same vehicle if it has previously been scanned by the system, or it utilises a reference image of the same model of the vehicle in the system database. In the latter case, the reference image is recovered from the reference database based solely on its undercarriage match via the Gatekeeper pattern recognition algorithm. To perform this comparison effectively, the imagery should be consistent from scan to scan, both day and night, to allow for accurate change analysis of the imagery.
The Gatekeeper IVUS uses many specific technical approaches in the design to provide consistent imagery and performance in all environmental conditions while other systems can struggle to produce effective results.
About Rapiscan
Rapiscan Systems (www.rapiscan-ase.com)| AS&E is part of the OSI Systems family of security companies. We provide cargo and vehicle inspection systems and services to help borders, ports, military, high-threat locations and law enforcement combat terrorism, smuggling and trade fraud. Our broad array of scanning systems leverages optical inspection, X-ray inspection, and radiation detection technologies to address each customer’s unique security application requirements.
Customers can further enhance inspection capabilities and operational efficiencies with our analyst assist tools and data integration platform. With this technology, information can be collected and combined from multiple sources in the inspection operation to automate manual processes, control workflows and deliver targeted, actionable intelligence. Our commitment to excellence in imaging performance, cutting-edge digital technology and uncompromising quality ensures our solutions help customers find threats and contraband with ease and confidence.
Parallax: the big problem Gatekeeper technology solves
Parallax is the apparent change in the location of objects as viewed from different positions. If a UVIS is not designed correctly, it will produce inconsistent imagery from parallax if a vehicle is not centred identically every time it passes over the scanning system. Just a couple of inches of movement of the vehicle to the left or right when passing over the sensor can result in very large changes in the apparent location of components in the imagery of the underside of the vehicle.
Since foreign object detection relies on change detection, systems that suffer from parallax problems are less likely to reliably detect foreign objects. Parallax problems can arise in a UVIS that has a small optical aperture located at the centre of the vehicle. To combat this problem, physical guide rails or curbs are often installed in the middle of the lane to force vehicles into the same location each time they pass over the sensor. Physical guide rails and constraints are understandably undesirable solutions for many customers given that vehicle tyres, undercarriages, and trailers can be potentially damaged as vehicles proceed down the lane. In addition, guide rails are not always effective solutions for centering vehicles with different wheelbases.
Failure to correctly align vehicles in systems with parallax problems could mean the imagery produced is significantly different, making a change analysis algorithm incorrectly identify these changes as foreign objects.
Similarly, objects at the edge of the vehicle’s undercarriage can be obscured from view due to line-of-sight blind spots from imaging from the centre of the vehicle outward rather than imaging directly upwards. The patented Gatekeeper IVUS design avoids these problems and allows for imaging without line-of-site or parallax issues.
Two views for twice the protection
The Gatekeeper-patented dual-view imaging system views the undercarriage from two different angles to identify objects that are hidden above pipes, mounts or other undercarriage infrastructure. This unique capability can help to identify more contraband than other UVIS. Line-scan cameras also do not have the ability to produce more than a single view of the undercarriage which limits their ability to observe the vehicle from multiple angles for maximum foreign object detection performance.
Change analysis requires consistent imagery, and a major component of producing consistent imagery is the uniformity of the illumination. Since foreign object detection is based on change detection, changes in illumination between the most recent vehicle scan and the reference scan can be identified as foreign objects. For example, significant changes in shadows or illumination brightness between the current vehicle image and the reference vehicle image can result in false detections of foreign objects and make manual inspection of the imagery by a human difficult. The Gatekeeper IVUS does not suffer from the same illumination non-uniformity issues as some competitors. Both the image brightness and uniformity are consistent day and night, allowing for accurate change detection when comparing new vehicle images with reference images, regardless of the time of day.
In some systems, it is possible that the difference in daytime and nighttime imagery is significant, making change analysis and foreign object detection difficult, if not impossible, due to the changes in illumination brightness and uniformity.
Image consistency: full-colour imagery is not an advantage
The Gatekeeper IVUS creates monochrome imagery for a specific purpose: to create consistent imagery for the vehicle’s undercarriage in all environmental conditions and lighting. To produce the consistent imagery needed for accurate change detection, the system must reject the sun’s impact on the amount of light collected by the camera; otherwise, there will be large changes in the data from scan to scan.
The Gatekeeper IVUS incorporates spectral filtration matched with its illumination to reject all wavelengths of light other than what is beneficial to the system. This spectral filtration creates a monochrome image that is ideal for change analysis, and it offers performance superior to colour imagery systems that inherently cannot use spectral filtration to improve image consistency. The monochrome imagery produced by the Gatekeeper IVUS has more consistent brightness and contrast, producing more accurate results with change detection algorithms, and is much more interpretable by humans during manual image inspection.
Is it really foreign object detection?
Some providers claim to have foreign object detection but instead use magnet detection on the underside of a vehicle. While the vehicle traverses the imaging system, the magnetic field is measured under the vehicle, which searches for magnets used to secure foreign objects to the bottom of the vehicle during a demonstration.
The magnetic detection software identifies the location of the magnet rather than the foreign object itself. Without the presence of a magnet, the object is unlikely to be detected under the vehicle if it was secured to the underside by other common methods (adhesive, bracketry, etc.). Magnetic detection technology also would not be able to identify other physical changes made to an undercarriage to mask the carrying of contraband.
Imaging system performance: line-scan vs. area-scan
The Gatekeeper IVUS uses area-scan imaging rather than line-scan cameras used in other systems, as area-scan technology produces an undistorted image of the vehicle undercarriage regardless of the vehicle’s motion over the sensor. Even if the vehicle speed changes while it’s traversing the system, or the vehicle stops before proceeding, area-scan imaging produces undistorted imagery by acquiring overlapping two-dimensional imagery at high frame rates and stitching them into a single, high-resolution image.
This technology enables a more accurate representation of the undercarriage without the blurring or distortion that is typically observed in line-scan camera imaging solutions that produce a one-dimensional image and attempt to create an accurate two-dimensional image from that data. The line-scan imaging systems are also sensitive to vehicle speed and driver inconsistencies that can result in distorted or blurry images.
Test who is best
Gatekeeper imaging systems from Rapiscan | AS&E are leaders in intelligent optical inspection technology, trusted by customers around with world with thousands of installations at borders, ports, critical infrastructure and highthreat facilities. We encourage on-site, side-by-side demonstrations with competing systems to demonstrate how our technology can more effectively help customers solve the most demanding security challenges.
