DocumentCode :
2356504
Title :
Combination of GPR and metal detector for mine detection
Author :
Jägerbro, P. ; Brusmark, B. ; Christiansen, A.-L. ; Lauberts, A.
Author_Institution :
FOA Defence Res. Establ., Sweden
fYear :
1998
fDate :
12-14 Oct 1998
Firstpage :
177
Lastpage :
181
Abstract :
This paper reports the progress of a three-year project with the objective to demonstrate multisensor techniques for mine surveillance. Data fusion techniques are using ground penetrating radar (GPR) of type ERA and a metal detector (MD) with induction coil of type Schiebel. The integration with a biosensor has been postponed until further technical development. The data fusion process runs in three main steps. First all sensor reports are converted to likelihood´s according to type. The likelihood´s are based on a priori test measurements with buried objects of different materials and sizes (low metal content mine, metal shrapnel, wax, stone, sand). In the second step the joint sensor report during a given time period, e.g. one measurement with MD/GPR, is computed as a product of probabilities. Finally, nearby measurement points are updated along the scanning MD/GPR system using a Bayesian fusing method. Compared with single sensors, the two-sensor system results in enhanced probability of detecting an object with both metal content and a surrounding shell of contrasting material, e.g. a low-metal content anti-personnel mine. At the same time the false alarm rate is significantly lowered
Keywords :
buried object detection; Bayesian fusing method; a priori test measurements; buried objects; data fusion techniques; enhanced probability of detection; false alarm rate; ground penetrating radar; impulse radar; low-metal content anti-personnel mine; man-portable mine detector; metal detector; mine detection; mine surveillance; multisensor techniques; two-sensor system;
fLanguage :
English
Publisher :
iet
Conference_Titel :
Detection of Abandoned Land Mines, 1998. Second International Conference on the (Conf. Publ. No. 458)
Conference_Location :
Edinburgh
ISSN :
0537-9989
Print_ISBN :
0-85296-711-X
Type :
conf
DOI :
10.1049/cp:19980715
Filename :
731296
Link To Document :
بازگشت