In attacking military and civilian targets, terrorists use many types of weapons, including car and truck bombs. Other potential vehicles for bombs include shipping containers and large packages or parcels. X-ray imaging presents the most effective method for detecting the contents of a vehicle, but developing x-ray systems large enough for such screening presents a challenge. Fluoroscopy Versus Radiography
There are two basic types of x-ray equipment capable of screening packages: fluoroscopic systems and radiographic systems. Both use a phosphor coating applied to a more-or-less rigid substrate to make the x-ray image visible. Fluorescent coatings glow immediately at visible wavelengths when struck by x-ray photons. In radiography, a luminescent coating stores energy absorbed from x-ray photons by exciting electrons into metastable quantum states, which can be read out later to make a visible image. Fluoroscopy, therefore, provides a visible image in real time but leaves no permanent record, while radiography provides a more-or-less permanent record, but requires a second step (development) to recover it.
Fluoroscopy is the workhorse x-ray technology for security applications. All of the airport x-ray equipment used to probe carry-on items, and more recently entire suitcases, are fluoroscopic systems mounted in an enclosure with a conveyor that carries the objects under inspection between the x-ray source and the fluorescent screen. Modern fluoroscopic systems use a video camera to pipe the visible image from the fluorescent screen to a television monitor, thus keeping operators and bystanders protected from x-ray exposure.
In contrast, medical applications rely almost exclusively on radiography. Radiographers have largely replaced the old photochemical emulsions with photoluminescent coatings. The electrons in these coatings can remain in metastable excited states for as long as eight hours at room temperature. "Developing" the image requires re-exciting these electrons back to the material's conduction band with photons in the material's excitation wavelength band, whence they imme-diately drop back to the ground state by emitting visible-light photons. A photodetector can capture the emitted light for subsequent image reconstruction.
Radiographic plates have an advantage: Technicians can "erase" them by bathing them in relatively bright light at an excitation wavelength long enough to ensure that all electrons return to the ground state. Once erased, the plates are ready for re-use. The process can be repeated indefinitely. The advantage for security applications is that technicians need only carry a limited supply of plates into the field, where they can re-use them.
The size of a fluoroscopic system is directly related to the size of the object it is screening. The fluoroscopic systems used at airport security gates typically feature tunnel widths of 18 to 36 in. (46 to 92 cm) and cost anywhere from $25,000 to $150,000 depending on size and x-ray power source, and other features.
These systems clearly cannot be used in many of the most critical security applications. Although it may be technically possible to scale them up large enough to inspect a vehicle, such a system would be very expensive, would require weeks to set up and take down, and would need considerable power. It could not be used to set up a checkpoint along a back road leading to Baghdad, for example. For a viable alternative, we must turn to radiography. Practical Imaging
Radiographic imaging using standard lumi-nescence-based plates developed for medical applications provides a much more feasible solution to the inspection problem. Radiography plates, which are essentially luminescent coatings applied to plastic substrates, are larger, lighter, and (by virtue of being mass produced) much less expensive. A standard-format x-ray plate is 14 * 17 in. (356 * 432 mm) and can weigh just ounces.
To image the back of a sports utility vehicle, for example, security forces would arrange multiple plates to form a large screen on one side of the vehicle. To expose the plates, a technician would move a portable x-ray source in a pattern on the opposite side of the vehicle. The whole exposure process would take just a few minutes using a battery-powered version of the prototype equip-ment, which could fit into a soldier's backpack.
The challenge has been to develop a means to read out the latent images captured on the plates. Conventional readers are typically the size of a large desktop copier and cost about $60,000, so they cannot be deployed easily to a large number of security forces at one time. They are certainly not man-portablethey must be airlifted by plane or helicopter to the task area along with adequate power-generation equipment. This procedure is logistically complicated and costly, both in terms of time and money. Making radiography practical has required the development of an economical scanner technology that could be made man-portable, battery-powered, and robust enough for field deployment under difficult conditions. Building the Perfect Beast
We started with a laser-based readout system. The phosphor on the radiographic plates is a mixture of three different barium fluorohalides (BaFI:Eu2+, BaFCl:Eu2+, and BaFBr:Eu2+) doped with europium as an activator. The wavelength band needed to stimulate luminescence from excited europium atoms is 355 nm, while the main emission appears at about 615 nm. Our system incorporates a diode laser operating at 670 nm (Edmund Industrial Optics; Barrington, NJ) matched to the phosphor's excitation band of 630 to 670 nm. It also includes a photomultiplier tube (PMT) and a fiber-optic collection system to gather the luminescent glow for conversion into an image.
The practical implementation of such a system presented a number of design issues, including the mechanical challenge of rastering the beam spot fast enough to read out the entire plate within a reasonable amount of time and the optical challenge of keeping the laser beam focused as it passes over the plate.
To scan a 432-mm-long plate in 30 s at 150 dpi (0.17 mm) resolution requires approximately 2541 scans at 85 Hz. It would be impossible to achieve reasonable scan rates by translating the laser or the filmthe masses are simply too large to achieve reciprocating motion over the distances required with any reasonable repeat rate. Beam-steering mechanisms based on electromechanically actuated tilting mirrors such as galvo scanners have achieved higher rastering rates than translating mechanisms, but the rates are not high enough when scaled to the beam sizes and maximum deflection angles required. Only a rotating mirror can achieve the required performance.
Multifaceted polygonal mirrors reduce the required rotational speed by a factor equal to the number of facets (see figure 1). To achieve an 85-Hz raster rate with a 10-facet polygonal mirror requires a rotational speed of only 510 RPM, as opposed to 5100 RPM for a single-faceted mirror. To achieve the required performance, scan angles must be large to keep the readout scanner's dimensions reasonable; the maximum beam-deflection angle typically exceeds 45°.
Figure 1. In the laser readout system, a rotating 10-facet polygonal mirror deflects the beam through an arc so that the beam spot traverses a line parallel to the y-axis 10 times per mirror rotation. Astigmatic aspheric optics flatten the field; meanwhile, a roller system feeds the x-ray plate through the machine at a constant speed. An all-reflective optical design with a folded path allows the system to fit into a compact package.
Dealing with the optical issues is even more challenging. Beam focus varies in the scan direction, with beam-path length varying inversely as the cosine of the arctangent of y/h, where h is the distance from the moving mirror to the film center. The system must therefore focus the beam more closely when aiming near the center than out at the edges. Optical magnification must be constant because variations in spot size will change the effective resolution. As the focal length varies with the y coordinate, the magni-fication has to be corrected to keep the spot size constant.
The design must also compensate for diffraction effects. Because both the dimensions and the deflection angles are large, diffraction effects can become significant, varying appreciably with y. Thus, the optical design needs to compensate for diffraction effects as well as optical aberrations.
The system uses an astigmatic, aspheric reflective-optics system to compensate for diffraction and optical aberrations while folding the beam path into a compact package. Image distance and magnification vary as a function of the deflection angle such that the spot size and shape does not change as the beam traverses the plate width.
We designed our prototype system to scan a standard radiographic plate. A photomultiplier tube (PMT) and a fiberoptic collection system gather the luminescent glow. Signal conditioning and digitizing electronics digitize the PMT's output and send it to an external laptop computer, which reconstructs the image and stores it in digital form. The imager can digitize a standard CR plate at 150 dpi in less than 30 s.
The existing prototype system is small enough to fit into a soldier's backpack. The current version runs on 120 V AC power, but another version is being built to run from a standard 12 V battery pack. To use the latter field version, a technician would need to carry the scanner, a small supply of CR plates, a battery pack, and a laptop computer.
Laser readout with a rotating polygon scanner can make x-ray inspection practical and portable for field use. With careful engineering, optical technology can surmount the design challenges in security applications and provide potential starting points for future systems. oe
George Murray is marketing manager of Axsys Technologies Imaging Systems, Rochester Hills, MI.