Pol-SAR Image Quality Enhancement Algorithom Based On The Scattering Model And NLM Filter

In the modern national economy, especially in the field of military action, polarization synthetic aperture radar(SAR) technique is a kind of advanced remote sensing technology, it is more and more important. Polarization SAR technology not only inherited the synthetic aperture radar technology throughout the day, all-weather, the advantages of high resolution, and compared with synthetic aperture radar technology polarization SAR can obtain a variety of echo signal, which can provide a more comprehensive scattering information. However, polarization speckle noises in SAR image make the quality of SAR image information is greatly reduced, and the existing polarization SAR image quality enhancement technique in noise reduction and not well maintain the polarization scattering properties of SAR image, in the polarization of SAR image classification, segmentation, target detection in the application of such have also been affected. As a result, the quality of the polarization SAR image enhancement technique is important in the study of polarization SAR image correlation processing, at the same time, in the process of long-term research found that master the polarization scattering characteristic similarity between SAR image pixel is more important, the polarization SAR image quality enhancement, classification and target detection is of great significance to study, however, the existing method of similarity measure the characteristics of scattering not well to measure the polarization scattering properties of SAR data. Polarization SAR based on synthetic and real data statistical analysis, put forward a new method of similarity measure the polarization characteristics, and its application to the quality of the polarization SAR image enhancement technique study,the main work is as follows:(1) Based on a large number of synthetic and real polarization SAR data characteristics of research, puts forward a new method to measure similarity between two polarization SAR. The measurement of polarization SAR pixels first polarization target decomposed based on scattering model characteristic vector, will get a vector to establish three-dimensional spatial polarization characteristics, at the same time, use of space Angle between the characteristic vector modulus and their relationship of two pixel similarity, compared with the threshold to determine whether two pixels.Through the validation test found that the method can accurately and is simple to measure thesimilarity of two pixels.(2) Polarization SAR image quality enhancement algorithm based on the scattering properties, we will polarization similarity measure method is applied to the SAR image quality enhancement algorithm study, put forward two methods of similarity measurement based on polarization polarization SAR image quality enhancement algorithm. The first kind of algorithm for each stay filtering pixel data type classification, determine whether low scattering energy points, whether has the scattering characteristics of aggregation, and according to the data types used to measure similarity between two polarization method to find its corresponding collection of similar pixels, and use the Lee filter model is coherent spot check. The second algorithm is the same for each stay filtering data types pixels are classified, whether pixels with scattering properties aggregation, and according to the data type by using polarization similarity measure method to find its corresponding collection of similar pixel, then,use the NL- Lee filter algorithm to calculate the parameters to increase the use of NL- Lee filter algorithm to calculate the parameters k(i, j) and the use of pixels after isolation of judgment to calculate the weight values w(i, j) and to get the final filter algorithm.