Point By Point Scanning Type Of Millimeter Wave Radiation Imaging Algorithms

Millimeter-wave (MMW) radiometric imaging has offered great potential in a wide range of military and civil practical applications, such as remote sensing, blind landing, navigation and security scanning. However, due to the confinement of the antenna aperture size and other reasons, the millimeter-wave radiometric images possess low spatial resolution. In view of the problem above, this dissertation focuses on the theoretical and experimental study of several key problems, including improving the detection imaging performance of the system itself, noise reduction and image recovery of the millimeter-wave radiometric image, edge detection of the millimeter-wave radiometric image based on noise reduction and image recovery process. The detailed contents are presented as follows:(1) The fundament theory of millimeter-wave radiometric measurement is analyzed; Different millimeter-wave radiometric imaging systems are introduced, and then, the hardware and software system structure of millimeter-wave radiometric imaging system of this dissertation are given.(2) According to the point-by-point scanning type of millimeter-wave imaging system, this paper presents a deep analysis of the related factors which affect the quality of millimeter-wave radiometric images, including the selection of the scanning mode, the spatial resolution and the sampling interval, sensitivity, the integration time and the scanning speed; The uncertainty equation suitable for point-by-point scanning system is derived, and it provides theoretical basis for setting best scanning and imaging parameters according to the application background; According to the problem of motion-blur imposed by the point-by-point scanning type of millimeter-wave imaging system, a novel motion-blur length and direction identification method which is based on Radon transform and differential autocorrelation is proposed and realized.(3) Several common standards which is used to measure the image quality are summarized; According to the de-noise processing of millimeter-wave radiometric image, by optimizing the existing algorithms, proposed wavelet adaptive threshold de-noising algorithm and de-noising algorithm which based on mathematical morphology and wiener filter. Then the above algorithms are used for the millimeter-wave radiometric images de-noise processing, the results verify the feasibility and effectiveness of the algorithm; Reasons of the millimeter-wave radiometric image degradation are analyzed. The image degradation model of the millimeter-wave radiometric imaging system is given. Image restoration is completed with the Wiener filter, according to the estimation result of the motion-blur length and direction.(4) Image edge detection algorithm which based on differential operator and image segmentation algorithm which based on watershed are realized, and the feasibility and effectiveness of the algorithms are verified. The pre-processing software of the millimeter-wave radiometric imaging system is given, which create conditions for the practical of the point-by-point scanning millimeter wave imaging system.