The system-development requirements will drive the system design. Each requirement needs to be analyzed and validated to ensure that valid trade-offs between performance requirements and system design issues can be made.2

A comprehensive discussion of the screening technologies in various environments (indoor and outdoor, controlled and uncontrolled areas, day and night, and so on) and for various scenarios (pat-down search, surveillance, tracking, and so on) is found in a National Institute of Justice Guide 602-00, Guide to the Technologies of Concealed Weapon and Contraband Imaging and Detection.3

The goal of any millimeter-wavelength/terahertz imaging system will be not only to locate an object of concern but also to identify what it is. This process of identification begins with detection and progresses through processes variously described as recognition and classification. For this report, “detection” is defined as “the process for discriminating objects of possible interest from their surroundings.” An operator, however, may not know what type of object is detected but only that something was detected. Conventional walk-through metal detectors will let the operator know that metal objects have passed through the portal, but these systems do not provide the location or the identification of the type (gun versus keys) of the metal objects.

The next level of sophistication is to acquire images of the detection space and then to use image-recognition algorithms to convert the image into an indication (such as an audible or visual alarm). The imaging and image-recognition capability requires access to, or possession of, a large information (data) storage capability and significant computing power to provide the real-time detection capability of finding contraband hidden on individuals in a line of moving people.

The recognition process must follow a strict hierarchy of algorithms with ever-increasing thresholds in order to arrive at a positive indication with a high probability of recognition and low occurrence of false recognition. The algorithm taxonomy begins with the detection of an item or object of interest, followed by a decision on classification as threat or nonthreat. An item classified as threatening is further examined in order to recognize the threat, for example, a weapon or firearm. A package of explosives may be recognized as an anomaly in the body image because the reflective properties of the explosive differ from the reflective properties of a human body, as shown in Table 4-1. A further refinement is identification. The identification step may be necessary in order to reduce false positives generated by prosthetics, shoe shanks, and so on.


R.J. Hwu and D.L. Woolard, eds. 2003. Terahertz for military and security applications. Proceedings of SPIE [International Society for Optical Engineering], Vol. 5070; P.H. Siegel. 2002. Terahertz technology. IEEE Microwave Theory and Techniques 50(3): 910; E.R. Muller. 2003. Terahertz radiation: Applications and sources. The Industrial Physicist, August/September, p. 27; D.M. Mittleman, M. Gupta, R. Neelamani, R.G. Baraniuk, J.V. Rudd, and M. Koch. 1999. Recent advances in terahertz imaging. Applied Physics B: Lasers and Optics 68(6): 1085-1094.


National Institute of Justice (NIJ). 2001. Guide to the Technologies of Concealed Weapon and Contraband Imaging and Detection. NIJ Guide 602-00. Prepared for NIJ, Office of Science and Technology, by Nicholas G. Paulter, Electricity Division, National Institute of Standards and Technology. Washington, D.C. February.

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement