Design and Implementation of Quadrotor Guidance and Detection System Hardware for Passing Through Window Based on Machine Vision

Quadrotor is one of the types of flying robots that has been highly regarded by researchers due to its simple structure and vertical flight capability. This paper presents a new method based on machine vision for the correct detection of window in an ever-unknown environment. One of the challenges of controlling the path of a quadrotor in an unknown environment is to accurately identify the window through which it passes. In this research, Parrot Bebop2 quadcopter is used, which is equipped with a camera. Also, an algorithm is proposed to perform image processing to identify the window in the environment and control the movement path of the quadrotor, which is implemented on the quadrotor. This method consists of three parts: processing, detection and identification. First, by applying image processing algorithms, we improve the image and delete data unrelated to the intended purpose. In addition, to control the path of the quadrotor, a proportional-integral-derivative controller has been designed and implemented using the Ziegler and Nichols method, which will be performed during a real indoor flight and in an automatic path tracking. According to the obtained results, it can be concluded that the use of flying robots can have positive results in military processes and low delivery to people in a short time.