| Popis: |
Landmine detection is a hazardous operation. To reduce the human casualties during landmine detection, unmanned ground vehicles (UGV) can be used. There are many landmine detection modalities that can be used with a UGV. However, the ground-penetrating radar (GPR) and the metal detector (MD; electromagnetic induction sensor) are fast and may be the most reliable means of landmine detection. In addition, GPR and MD sensors can easily be packaged in one sensor head. It is also easy to configure the sensor heads in an array. To combine GPR and MD sensor data, a series of experiments has been performed. The GPR and MD combined sensor is scanned over a 2 m × 2 m × 1 m soil box at a constant height. The soil box contains stony sand and a landmine, which is buried at 200 mm depth when measured from the surface of the ground to the top of the landmine. The control parameters for the experiments are sensor height and soil moisture. The sensor height is varied from 40 mm to 100 mm with 20 mm increment from the soil. The soil moisture conditions considered in the experiments are dry and wet. The stony sand used in the experiments contains rocks with approximately 100 mm diameter scattered in a random manner and a brick placed 200 mm apart horizontally from the landmine. The GPR and MD responses are recorded for all control parameter combinations and later analyzed. The GPR images are generated using time-domain back projection (TDBP) algorithm. The GPR images and MD responses are compared for the soil conditions and sensor heights. A combination algorithm, which enhances GPR images using MD responses, is also presented. |
