Research On Application Of Specific Target Rfcognition Technology For Aerial Image Of Uav Based On FPGA

In the field of image research,the identification and real-time tracking of specific targets in images has been the technology that many scholars have continued to study deeply and need to be further overcome.Image recognition and real-time tracking technology in the UAV has been widely used in aerial photography,such as earthquake investigation,military reconnaissance,or investigation,engineering mapping,crop yield and so on.Image-specific target recognition and tracking to achieve the realization of a certain degree of hardware and software to be achieved,usually for most of the ordinary image processing can be done through the software,but in the process of military UAV aircraft generally in high-speed operation,real-time High-resolution shooting situation,only rely on software for image processing has been unable to meet the actual needs,so we need to combine the ARM processor scheduling and FPGA hardware acceleration function to obtain real-time image analysis to achieve a specific goal,Attack target identification and real-time tracking and other functions.With the popularization and application of embedded system,the cost of image recognition real-time tracking system is declining,and the hardware is also going to be miniaturized and the power consumption is decreasing.Therefore,FPGA-based UAV aerial image special target recognition technology Application research to choose the use of hardware to achieve.In order to use the soft and hard core co-processing real-time analysis of the specific objectives of the acquisition and judgment,we propose the system's hardware mainly includes the following IP core:DSP digital signal processor,integrated graphics processing chip or FPGA semi-custom logic circuit with other Auxiliary logic circuit,compared with the first two hardware design,the use of FPGA semi-custom logic circuit with more hardware resources can choose,with low cost,good performance and stability characteristics,while FPGA easy to maintain the maintenance can provide more convenient Of the upgrade to achieve functional enhancements,so that the development of UAV aerial image specific target recognition system will have greater versatility and elasticity index.In this paper,the background and significance of the research are introduced in the first chapter,and the latest research results of image recognition and real-time tracking technology,UAV aerial technology and FPGA technology are reviewed.In chapter 2,the basic constitution of the specific target recognition method system for aerial images is expounded,the particularity of UAV aerial photography and the advantages of aerial image processing with FPGA are expounded,which lays the foundation for the development of the follow-up system and provides the theoretical basis.In the third chapter and the fourth chapter,the hardware and software of the target real-time recognition system of the UAV aerial image based on FPGA are designed and elaborated.In the fifth chapter,the framework of UAV aerial target recognition process is introduced,and the experimental system platform is built.The edge detection,active contour model,Sobel operator and Snake model are combined for the specific target real-time recognition system of UAV aerial image Contour extraction experiment,through the experiment proved that the system involved in this study to achieve a specific target recognition function.