Research On ZYNQ-based Dynamic Target Image Acquisition Technology

In order to realize the automation and unmanned operation process,the train formation station or automatic loading platform needs to measure the speed of slow trains,which is the only way to coordinate with the equipment and traction locomotive action according to the speed of the car.The basic principles of optical flow speed measurement are explained in depth,and the requirements of slow train speed measurement are analyzed,especially the speed measurement accuracy and dynamic responsiveness are explored in detail,and a ZYNQ-based slow train speed measurement system is applied to solve the problems of high speed measurement accuracy and strict integration requirements in the process of train speed measurement.This paper designs a slow train speed measurement system based on ZYNQ,and the core processing chip of the system is Xilinx’s Zynq-7000 SOC series;then the SCCB protocol is used to control the camera,and the AXI4-Stream protocol is used to move the data to achieve realtime image data acquisition and transmission;then for the internal FPGA Then,to address the problem of small storage capacity inside the FPGA,DDR3 SDRAM external memory model NT5CC256M16 is used to complete the real-time caching of image data,and the individual storage capacity of this chip is 4Gbit,which meets the requirements of this design;to address the problem of lack of serial port in the upper computer,CH340 C chip is used to realize the image data transmission to the upper computer;finally,the acquired image data is displayed in real time on the LCD through HDMI transmission interface.The acquired image data is displayed on the LCD screen in real time through the HDMI transmission interface,while the image data processed by the optical flow method calculation module is transmitted to the host computer through the serial port.ZYNQ-based slow train speed measurement system is established,which mainly includes camera acquisition module,optical flow method calculation module,data caching module and data transmission module.For the optical flow calculation module,the traditional optical flow method is analyzed and improved so that it can be implemented in ZYNQ platform;for the data caching module,the storage space of DDR3 SDRAM is appropriately divided and three frame buffer mechanisms are used for image data processing to complete the uninterrupted caching of moving target image data.An experimental platform is built for testing,and the stability of image data transmission and optical flow calculation algorithm of ZYNQ-based cache slow train speed measurement system is verified for several times.After a certain number of sets of repeated tests,through the analysis and summary of multiple sets of test results,it shows that the system is suitable for slow moving target ground speed analysis under the conditions of resolution 1280x720 and moving target speed in 0-30cm/s.