At present there are approximately 110 million landmines scattered around the world in 64 countries. Now days, the clearance of these anti-personnel mines takes place manually. Unfortunately, an average for every 5000 mines cleared one mine clearer is killed. So, needs for automatically anti-personnel mine detection system have been increased. Many anti-personnel mine detection methods are under development, but there is no mine detection system, which finds anti-personnel mine perfectly. So, an anti-personnel mine detection system by contact force is proposed in this research. Contact force profile is affected by the material, mass, and shape of contact bodies, and contact velocity. When impact velocity and the shape of impact bodies are fixed, contact force profile is the function of the material and mass of impact bodies. So we can estimate the material and mass of impact bodies in this condition. Contact force measurement system applicable to anti-personnel mine detection system is proposed and contact force profile is characterized by four terms, which are used in learning process. Radial Basis Function Network (RBFN) is used to estimate the mass and classify the material of body. Mass estimation and material classification RBFN is proposed and evaluate the reliance of the RBFN. And finally, evaluate the possibility of the application to anti-personnel mine detection system.
