The Design Of UAV Visual Stabilization Platform System

In recent years, unmanned air vehicle (UAV) in civil, industrial and military fieldsreceived the widespread attention. With the rapid development of computer technology,automatic control technology, image processing technology and so on, the ground targettracking UAV who carried vision tracking device comes into being and receives more andmore attentions. Its major function is capturing moving objects, automatically tracking thetarget and real-time image transmission to the master control platform. So it can be used forthe precision guidance, Counter-terrorism, road-cruise, emergency and disaster relief and soon.In this paper, we use a small four-rotor UAV as a carrier, and select the CCD camerawhich installed on a dual-axis stabilized platform to capture video images. The platform cannot only isolate the disturbance of UAV, in order to ensure stability of the camera to obtainimages and make the camera track the target in response to control signals. This design is avision tracking systems, including multi-disciplinary knowledges such as stability control andimage processing. Because of the complex application environment and huge outsideinterference, the design of software, hardware and tracking control methods have greatlychallenge.In order to achieve tracking rapidly and steadily for moving target, a series of questionsare researched in this paper.Firstly, understand the background and significance of the reseach about the UAV visionpan-tilt head. According to the system’s coordinate and the principle of dynamics infer thecompensation formula of isolation carrier’s disturbance, and establish system’s mathematicalmodel. After analysis in frequency domain, make anticipatory control on stable rings andcontrast simulation, I choose the fuzzy PID controller. Analysis of factors affecting theaccuracy of the system, design the algorithm to compensate the gyro drift. Then, according tothis system’s application environment, analyze and classify the existing tracking algorithmicand select several types of appropriate algorithms. Hardware diagram of system is presented.For dual-axis head frame, select the appropriate electronic devices and peripheral devices, anddebug control circuit board. Software diagram of system is presented. Knowing the collectionof sensor signal and processing methods, then compile various parts of code. Finally,according to system’s practical application and all kinds of disturbance from outside word,design the related experiment which test system’s stability and anti-jamming and accuracy. By many experiments to improve the system hardware, and confirm the different target trackingalgorithm’s advantages, disadvantages, I design appropriate algorithm.