Sar Image Resolution Enhancement Method

Synthetic Aperture Radar (SAR) is a microwave remote sensing imaging radar which can work day-and-night under all-weather conditions. With these abilities it has shown a great potential and application prospect in the fields of military and economic. These applications mostly depend on the quality of SAR images. The better the quality of SAR image is, the better of its application performance is. Image resolution is one of the most important standards to measure the quality. Because abundant information and obvious characteristics of targets can be extracted from SAR image with high resolution, many fields require a high level of SAR image resolution. Therefore it is very demanding and meaningful to improve the resolution of SAR image.In this paper, the regularization method and the super resolution construction method to achieve SAR image resolution enhancement are studied. The main contents are as follows:1. The background and significance of resolution enhancement are introduced, and the corresponding research situation and trends at home and abroad are summarized. Then the concept of resolution enhancement of SAR image is explained and the sparse distribution of SAR image is analyzed. Finally several image quality criteria are used to evaluate the image quality.2. Resolution enhancement of SAR image based on regularization theory is studied. The requirement for regularization method to be used in the area of SAR image is analysed and verified by simulation. And a method based on two-dimensional decomposition and regularization processing is proposed. This method is applied to SAR amplitude image, and a projection operator is set up between the SAR image and the real scene. The operator used here doesn't need as many system parameters as the operator does in frequency domain and it has two-dimensional orthogonal irrelevance, which provides the possibility to decompose the SAR image into two dimensions. Compared with the traditional SAR image regularization method, the choosing of the operator and the two-dimensional processing can reduce calculation. At last simulation results verify the ability of the regularization method to enhance image resolution and reduce noise.3. Resolution enhancement of SAR image using multiframe images is studied. Based on the analysis of the feasibility to use super resolution (SR) technology in the field of SAR image, the idea of SR is brought in the processing of SAR image. And an improved IBP (iterative back projection) method is proposed. In this method, at each iteration, a different coefficient is induced to obtain a new revised image, which makes full use of the high frequency information. Both simulation data and real data are adopted to verify the feasibility of the method, and results show that the image obtained by this method has more details and higher resolution. .