Detection of hidden explosives through fast-neutron transmission analysis of HCNOX content

Conference · Sat Dec 31 04:00:00 EST 1994

OSTI ID:199848

Overley, J C ^[1]

---

1. Univ. of Oregon, Eugene, OR (United States)

Computer simulations, based in many instances upon measurements, have been used to devise on algorithm for detection of explosives in luggage from projected number densities of H, C, N, and O. Other substances are lumped together as projection X. Luggage-thickness dependence is reduced by dividing each projection by the total projected number density. This normalization constrains projections to a 4-dimensional space. A distribution of normal materials (N) is built up by sorting results of some 100,000 simulations into a series of {open_quotes}rectangular{close_quotes} bins in that 4-space. Explosive distributions (B) are similarly established. A matrix of detection parameters is calculated from the ratio B/(N+B) for each bin. For a realistic distribution of the numbers and types of luggage materials, and under realistic laboratory conditions, the plastic explosive RDX, at 10% of suitcase thickness and in a single pixel, can be detected with 85% reliability with a false alarm rate of 3%.

• OSTI ID: 199848
• Report Number(s): CONF-941129--
• Country of Publication: United States
• Language: English

Related Subjects

45 MILITARY TECHNOLOGY, WEAPONRY, AND NATIONAL DEFENSE
AIRPORTS
CHEMICAL EXPLOSIVES
NEUTRONS
REMOTE SENSING
TRANSMISSION