The Design And Implementation Of Image-Based Runway Detection System

In this paper,a runway detection system based on image is designed and implemented for medium-large fixed-wing UAV,aiming at solving the problem of using camera to capture runway images to complete runway detection in the practical application of autonomous landing.Firstly,the research status at home and abroad is systematically discussed according to the research content,and the advantages and disadvantages of various embedded platforms currently used for image processing are pointed out.Under the comprehensive consideration of the requirements of runway detection performance,the overall implementation scheme of the runway detection system is designed.According to the visual landing process of medium and large fixed-wing UAVs,the runway detection system is composed of image acquisition unit and embedded image processing unit.The image preprocessing system implementation scheme and hardware equipment design scheme of runway detection system are designed.Secondly,the image preprocessing system of the runway detection system is realized according to the overall implementation scheme,and the image stream transmission channel and the preprocessing system are provided for the realization of the runway detection algorithm.The design process of the on-chip system of ARM+FPGA is introduced.Then,the PL system and the PS system based on ?C/OS-? are built respectively.An FPGA-based image preprocessing system with hardware acceleration is designed to realize the image streaming channel for real-time acquisition,transmission and display of images and the scaling of the image of the runway.Thirdly,the hardware design of the runway detection system is completed.The structure of the core board and the backplane of the hardware equipment of the runway detection system is designed,and the modules of the backplane are divided into the power module,the debugging module,the communication module,the data transmission module and the camera interface module.Then perform circuit design and PCB design according to these modules.Perform unit commissioning and board level debugging to provide hardware equipment for runway detection algorithms.Finally,a series of experiments are designed to verify the function of the runway detection system.An image streaming function test is designed to verify the image acquisition,transmission and display functions of the runway detection system.Then,verify the scaling function and hardware acceleration capability of the FPGA-based image preprocessing system.The runway detection function of the runway detection system is then verified to determine the runway area.Through a series of experiments,it is verified that the runway detection system designed in this paper has the ability to collect runway images in real time and realize runway detection.