Uav Monocular Machine Vision Drop Positioning Algorithm And Dsp Implementation,

Machine vision mainly simulates human visual function with the computer. It acquires information from the object images, and processes the information, and is used in detection, measurement and control finally.Machine vision researches how to enable the computer's ability of cognizing the 3D information through a two-dimensional image. The machines not only can sense the 3D geometric information, including its shape, location, posture, movement, but also on their description, collection, identification and understanding.In this paper, I mainly research how to obtain the attribute of the Unmanned Aerial Vehicle (UAV) including its posture angles and location parameters with the vision technology. For obtaining the attribute, I have to get some necessary information of the object images which are gained by the camera on UAV.UAV develops rapidly and uses widely after the Gulf War. The western countries recognize the UAV's function in war. They apply the high technology to UAV's development, for example the advanced signal processing and communication technology can raise images' digital transmission speed. Other advanced technical equipment, such as the advanced wiretapping, the radar that can penetrates the leaves and so on will also be installed on the UAV. UAV's reclaiming is very important for the UAV's flight safety. Therefore UAV's completely automatic landing technology is an research focus.UAV's posture angles and location parameters are necessary for its stabilized flight and the safe landing. The method of acquiring UAV's posture angles and location by machine vision for the realization of UAV's independent landing is of great significance. Because of deficient test condition and equipments, I research the images gathering by simulation software. I make detailed analysis to the connection between UAV's attribute and the horizon and runway's boundary straight line. And I get the solution of obtaining the UAV's roll angle and pitch angle from the horizon straight line parameters, as well as the yaw angle from the runway boundary straight line parameters. After obtaining UAV's posture angles I calculate the UAV's location parameters relative to the airplane main runway. Finally I realize the algorithm based on the DM642 hardware platform.This solution is low-cost, simple and practical. It is very valuable for the UAV's navigation and localization technology's development.