Research On Data Acquisition And Stripe Noise Suppression Method For Passive Millimeter Wave Imaging System

Passive millimeter wave(PMMW) imaging system achieves images based on the difference energy of millimeter wave radiation between the target scene and object. As to the system is non-radiative; can penetrate clothing, clouds, smokes; and is difficult to detected; the PMMW system has been widely used in the field of security inspection, military reconnaissance etc. In order to meet the demand of applications, PMMW systems always use multi-channel architecture to achieve real-time imaging. The data acquisition unit is the key hardware unit of the multi-channel imaging system, is responsible for the data collection of the channel output analog brightness temperature signal, in order to meet the imaging algorithm processing. However, limited by the technique of production process, the non-uniformity between multi-channel and multi-beam lead to stripe noise showing in the achieved image, the stripe noise badly affects the image quality.Based on the research of multi-channel PMMW system, this paper designs the multi-channel data acquisition unit based on the FPGA, and researches on the stripe noise suppression method. The main works are:1. Analysis the millimeter wave imaging theory and the multi-channel imaging system architecture. The main technical indicators of data acquisition are demonstrated and the least mean square error theory is studied, which can provide a theoretical basis for the hardware designing and the algorithm researching.2. Analysis and demonstrate the data acquisition unit structure and implementation scheme, complete the main device select and interface design. The “AD+FPGA” hardware structure is used to design and develop the multi-channel data acquisition unit, and complete the hardware circuit debugging.3. Using the Verilog language to design the sequential logic circuit on the Quartus software platform. The main functions of this module are implemented, such as data acquisition, data storage, data transmission. Complete the testing of data acquisition and transmission on the software platform.4. As to the stripe noise problem in the multi-channel imaging system. The method based on the calibration to equalize each channel is studied. By analyzing the multi-channel imaging system mathematic model and the characteristic of stripe noise, as to the low-pass effect of the point spread function of the imaging system, an improved moment matching method is applied. Compared with traditional moment matching method, this method not only has a good ability to stripe noise suppression, can also effectively eliminate the “band effect” in the denoising image.5. The algorithm based on the least mean square(LMS) is researched to destripe. According to the imaging mathematical model of the multi-channel system, based on the traditional LMS algorithm and taking the advantage of the line correlation and neighborhood related information of the image, an improved LMS algorithm is applied to the PMMW image destriping, the simulation results show that the method has good stripe noise suppression capacity, and retain the image details.By simulation and actual experimental, verify the data acquisition unit and destriping method researched in this paper achieve the desired requirements, which lay the technical and theoretical basis for the research of passive millimeter wave imaging system.