Multi Sensor Unmanned Aerial Vehicle Heading Control Based On FPGA Device

In recent years,the navigation control technology of UAV has developed rapidly,and the precise control of course is particularly important.Most of the indoor unmanned aerial vehicles(UAV)are mostly using wireless communication positioning inertial navigation and visual navigation technology,but the development of visual navigation is not mature and can not be widely used.Sexual navigation and GPS navigation technology are highly dependent on the surrounding environment,and are easily disturbed by environmental factors.In order to solve this problem,this paper builds a small four rotor unmanned aerial vehicle(UAV)as a flight platform according to the requirement of low altitude flight.It uses a single CMOS image sensor to study image processing algorithm of image navigation,enhance the image recognition of unmanned aerial vehicle,and solve the autonomous navigation of unmanned aerial vehicle under the disturbance of inertial sensors.The main contents of the design of UAV flight control and image navigation algorithm are as follows:First,it analyzes several common navigation methods of UAVs,points out the role of visual navigation in the course of modern UAV and its way of controlling course.It adopts the direction control method of multi sensor combined with visual navigation,and assembles the four rotor small UAV Based on the requirement of low altitude flying environment,which is based on the "X" character.The flight direction needs,using CMOS image sensor,combined with gyroscope and pressure sensor and other multi-sensor course control system,with FPGA device to realize the real-time and strong hardware processing system.Secondly,the image processing method of CMOS sensor is studied in detail.The two value processing method of the optimized improvement is proposed.The images are processed by gray scale and Gauss filtering.For the four rotor unmanned aerial vehicles(UAV)in the indoor and low altitude,the image processing method should pay attention to the thresholding of the image,and improve the image resolution and the image processing speed after the processing.It is very effective for large number of images to be processed in UAV navigation.Then,the multisensor processed by FPGA device is designed as a whole,which improves the speed of real-time image processing,makes the heading information more accurate and stable,and coordinates the sensor to each other,setting the image navigation method to navigate under the circumstances of the inertial sensor easily disturbed by the environment.Two best feature points are obtained by the feature matching method,which is processed by the previously processed images,and the heading angle is obtained by the method of the difference of the center of centroid slope.Finally,four rotor UAV is based on FPGA inertial sensor device navigation flight measurement experiment and improved image processing method of heading measurement simulation analysis closer to the actual condition in the inertial sensor data heading measurement based on FPGA device,heading angle measurement more accurate.The improved image processing method has good stability and accuracy,and it is feasible to combine with inertial sensors to achieve the desired purpose.