.3 Mm Band Signal Acquisition And Image Processing

At present, the Millimeter Wave(MMW) technology is getting more and more widespread in the military and daily life's application, for example, missile-mounted MMW radiometer guidance system, airborne and satellite passive MMW imaging system, airport security check system at the entrance and so on. To improve the application of the MMW technology, this paper conducts the research on 3 mm wave band signal's gathering and image signal's processing.In the signal gathering part, the paper first analyzes the basic principle of detection system. And then it makes one-dimensional detection experiments on some area targets such as water surface, lawn, pavement and metal plate by using the DC Dicke radiometer, from which we obtain the target's apparent temperature at different incidence angle and make a detailed analysis of the experimental data. Then, a two-dimensional detection experiment on scaled tank and metal plate is conducted by using the DC Dicke radiometer, with a comparative analysis of three-dimensional and plane objectives with the same material. Finally, we use the AC full power radiometer to take an image of the person of hiding metal objectives indoors for detecting metal objectives, and take an image of scaled tank outdoors.In the image processing part, the paper conducts the denoising research on the MMW image in spatial domain and wavelet domain. After analysis, the weighted gradient's reciprocal method, adaptive median filter and semi-soft wavelet threshold method are very good methods. In addition, to carry on later period's target identification well, this paper, in view of the MMW image characteristic, makes image enhancement processing by using image interpolation and histogram revision. And the simulation analyzes the effectiveness of these methods through Matlab, seeking for the most superior method. The experimental results prove that the cubic convolution method has the most accurate results, and the local area histogram equalization with interpolation of the histogram revision method is relatively good solutions.