FPCNN based Image processing for autonomous Quadcopter control

In this paper, an image processing technique based on FPCNN is proposed for autonomous Quadcopter landing. In the UAV field, take-off and landing are two important processes, as any little miss operation in each of them will cause a crash. In some cases, the operator needs to take-off and land automatically. Hence an automatic process can improve the efficiency of Quadcopters. An environment detection method, decision-making unit and actuators that control the unit are necessary for achieving this goal. In this paper, Image Processing and FPCNN (Implemented in MATLAB) are used for detecting the zone of landing, decisions for controlling the motor are made through an interface board which receives instructions from MATLAB and transfer them to the main controller board. By using this method, Quadcopters can land on the specified area, which is defined 1*1 m plate and there is H in the middle of it, without operator intervention. Details of the proposed method are presented in different sections. Results state the efficiency of the method.

Keywords:

Autonomus Quadcopter , FPCNN , Image Processing , landing , take-off