Research On System Integration And Channel Equalization For Passive Millimeter Wave Image

PMWW System is a system that utilizes the two-dimensional digital scene matrix with different values form from the millimeter-band energy of different objects in the scene,and normalizes the matrix to image.As PMWW System has the characteristics of safe and no radiation,non-contact inspection,easy to detect metal,penetrate textile clothing and so on,and the imaging results can realize the inspection of human body carrying hidden items,so this system so the system security at the airport,key regional monitoring and other fields have a wide range of prospects.With the improvement of performance requirements such as real-time and high-resolution in imaging systems,multi-channel imaging mechanism or array imaging mechanism is one of the future mainstream development trends.However,due to device discretization and manufacturing process and other factors,different channel response differences,resulting in imaging results appear band noise,the impact of subsequent image processing algorithms,so the need for channel equalization research.This thesis is based on the actual research projects,the specific work is as follows:Based on the basic theory of millimeter wave imaging,the architecture of millimeter wave passive imaging system is studied,and the passive detection of hidden metal objects is carried out.The channel equalization theory and related algorithms based on image domain are introduced.According to the requirements of the imaging system,the basic structure and implementation scheme of the data acquisition unit are studied.A local serial parallel data acquisition scheme is proposed,and the corresponding device selection and hardware circuit design and implementation are completed.Based on Intel FPGA development platform,Verilog HDL language is used to complete the digital logic design of data acquisition and control unit,which realizes the functional requirements of data acquisition,data storage and transmission control in this unit.According to the function and index of the servo scanning mechanism,the control method and the scanning structure of the servo scanning mechanism are studied.The realization scheme of the turntable for one-dimensional servo scanning is proposed and developed.Finally,combined with the display control subsystem,signal processing subsystem and the antenna front subsystem to complete the overall system integration test.Based on the actual imaging mechanism of the project,the basic channel equalization theory is studied.The improved method based on the moment matching and the improved method based on the neural network are proposed,and the band noise in the imaging result is suppressed or removed by these methods.