期刊名称：Archiwum Fotogrametrii, Kartografii i Teledetekcji
印刷版ISSN：2083-2214
出版年度：2008
卷号：18
语种：English
出版社：Main Board of Association of Polish Surveyors
摘要：Solving the problem of image smoothing is regarded as an essential stage in preparing digital images for further processing. It was tackled by a number of studies. The presence of speckle noise in SAR images is the major obstacle in interpreting,classifying,and analyzing SAR images. The main problem in many remote sensing applications is the extraction and interpretation of information about the objects which are present on SAR images. This makes the speckle noise reduction a very important task. The reduction of speckles was performed by applying the nonparametric method,described by J. Polsehl and V. Spokoiny in 2006;the method is based on locally constant smoothing with an adaptive choice for every pair of data points (Adaptive Weights Smoothing – AWS). The AWS algorithm has never been tested in detail on SAR data. This paper describes the methodology of using the AWS algorithm by integrating three images: one original image and two images determining the result of the algorithm processed. The performance of the proposed method was tested on high-resolution X-band synthetic aperture radar TerraSAR-X images and was compared with popular adaptive filters (Lee,Frost). The method presented was tested on two samples extracted from images captured in different imaging modes,with different geometric resolution and showing various land use and land cover. The results confirm the utility of the propagation-separation approach for radar image smoothing.
其他摘要：Redukcja szumów na obrazach jest istotnym etapem wstĊpnego przetwarzania danych. Zagadnienie redukcji szumów byáo wielokrotnie poruszane w licznych publikacjach naukowych. Przygotowanie danych przed wykonaniem wáaĞciwych analiz jest szczególnie waĪne w pr
