Evaluation of Test Results of GPR-based Anti-personnel Landmine Detection Systems Mounted on Robotic Vehicles

This paper discusses a data analysis and evaluation of test results for anti-personnel landmine detection systems using ground penetrating radar (GPR) mounted on robotic vehicles for humanitarian demining. Six research teams from universities and industries founded by Japan Science and Technology Agency has been developed GPR systems in combination with electromagnetic induction (EMI) sensors, or metal detectors (MDs). As well as sensor technology itself, highly-accurate sensor positioning using advanced robotics technology plays a part in providing operators with clear subsurface images. The concept of the developed systems is to make no explicit alarm and to leave decision-making using subsurface images to the operators like medical doctors can find cancer by reading CT images. To evaluate these kinds of systems, a series of tests have been conducted in Japan from 8 February to 11 March 2005. Since operators' pre-knowledge of the locations of buried targets significantly influences the detection result in the tests, all the 6 lanes are designed to be suitable for blind tests using more than 200 landmine surrogates. The test results showed that combining GPR with MD can improve probability of detection (PD) around a depth of 20cm, where it is difficult to detect targets by using only a metal detector and that there is a room for further improvement in the PD by feeding back the test results to testees to learn typical target images, where targets were not able to be detected in the blind tests. It has also been learned that positioning control must be improved in scanning the ground with a sensor head, which is a key to making the best of use of MDs mounted on vehicles.

[1]  L B Lusted,et al.  Radiographic applications of receiver operating characteristic (ROC) curves. , 1974, Radiology.

[2]  C A Roe,et al.  Statistical Comparison of Two ROC-curve Estimates Obtained from Partially-paired Datasets , 1998, Medical decision making : an international journal of the Society for Medical Decision Making.

[3]  K. Jayaraman,et al.  A statistical manual for forestry research. , 2000 .

[4]  S. M. Shrestha,et al.  High resolution image reconstruction by GPR using MUSIC and SAR processing method for landmine detection , 2003, IGARSS 2003. 2003 IEEE International Geoscience and Remote Sensing Symposium. Proceedings (IEEE Cat. No.03CH37477).

[5]  Motoyuki Sato,et al.  GPR using an array antenna for landmine detection , 2004 .

[6]  Yasuhisa Hasegawa,et al.  GPR-based adaptive sensing: GPR manipulation according to terrain configurations , 2004, 2004 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) (IEEE Cat. No.04CH37566).

[7]  Jun Ishikawa,et al.  Experimental design for test and evaluation of anti-personnel landmine detection based on vehicle-mounted GPR systems , 2005, SPIE Defense + Commercial Sensing.

[8]  Kensuke Takita,et al.  Teleoperated Buggy Vehicle and Weight Balanced Arm for Mechanization of Mine Detection and Clearance Tasks , 2005 .

[9]  Motoyuki Sato,et al.  Development of a hand-held GPR MD sensor system (ALIS) , 2005, SPIE Defense + Commercial Sensing.

[10]  Yasuhisa Hasegawa,et al.  Environment-Adaptive Antipersonnel Mine Detection System - Advanced Mine Sweeper , 2006 .