M. Mlynarczuk, S. Porzycka-Strzelczyk
Thursday 11 October 2018 by Libadmin2018


During the satellite SAR (Synthetic Aperture Radar) data analysis it is highly important to remove all the noise components from the image. The most critical disturbance that affects radar images is speckle noise. Many different filtration methods to reduce this noise have been proposed in the literature up to now. This paper presents the results of SAR image filtration using the methods of mathematical morphology. During this work over 20 different morphological filters have been tested – both classical ones (opening, closing, alternating filters, sequential alternating filters), and also less used morphological filters realized with using geodetic reconstruction. Different sizes of structural elements have been used. The tests have been done for TerraSAR-X SAR image acquired in the Staring Spotlight mode over the port in Singapore. For the analysis, the part of satellite radar image with a ship and surrounded sea surface was selected. Such scene is characterized by the homogeneity of the background (see) and sharp edges of the object (ship). The goal of the study was to check how each morphological filter deal both with equalizing the background and keeping the edges of the object. To study the homogeneity of the background, standard deviations of the selected parts of the see after filtration has been analyzed. The influence of the filtration on the edges has been assessed based on the analysis of the image profiles and sizes of morphological gradients on the edges of objects. Results show that the application of morphological operations can be very useful in SAR imagery filtration. By analyzing results of each filtration method it has been concluded that the best results have been generated by morphological filters that use reconstruction.

Keywords: speckle, SAR, morphological filters

Home | Contact | Site Map | Site statistics | Visitors : 0 / 353063

Follow site activity en  Follow site activity PHOTOGRAMMETRY AND REMOTE SENSING  Follow site activity Papers SGEM2018   ?

CrossRef Member    Indexed in ISI Web Of Knowledge   Indexed in ISI Web Of Knowledge

© Copyright 2001 International Multidisciplinary Scientific GeoConference & EXPO SGEM. All Rights Reserved.

Creative Commons License