Research on target recognition and path planning for EOD robot

In order to improve the autonomy of the explosive ordnance disposal (EOD) robot, the paper makes a research on the target recognition and path planning algorithm based on laser range finder and Time of Flight (TOF) camera. The target position of the explosive is selected by the operator from the grey-scale image, then EOD robots use Kernelised Correlation Filters (KCF) tracking algorithm and distance constraint to realise the real-time explosives identification. In the case of A* algorithm, it does not take the heading information of the robot into consideration, which may lead to the problem of heading shakes in the vicinity of the obstacle. In this paper, the heading constraint is introduced to improve the A* algorithm so as to eliminate the problem of heading shakes and improve the stability of the robot. The simulation and experiment results demonstrate the feasibility of the algorithm proposed in the paper.