Color identification for quadcopter flight control and object inspection

Abstract

This study focuses on quadcopter flight control and image application. The proposed quadcopter can be applied to inspect rust or damage areas of ship hull. Real-time image identification with proper devices is also proposed. Proportional–integral–derivative control and fuzzy control are used in flight control design. Due to hardware limitation of the quadcopter, complicate control theory is avoid in controller design. A simple fuzzy control is used in this study. In sensor fusion, air pressure altimeter, Global Positioning System (GPS), ultrasonic sensor, and optical flow sensor are utilized to maintain altitude holding in outdoor and indoor environments. In color identification, Pixy (CMUcam5) camera is used to perform image processing and image tracking. Signals from different sensors are processed in the Arduino board. In flight simulation, pitch and roll angle commands are constructed in a rule table. Flight control can be performed by feedback tracking error, and software simulations for proportional–integral–derivative and fuzzy control are given.