Title :
Prediction of Despeckling Efficiency of DCT-Based Filters Applied to SAR Images
Author :
Rubel, Oleksii S. ; Lukin, Vladimir V. ; De Medeiros, Fatima S.
Author_Institution :
Dept. of Transmitters, Receivers & Signal Process., Nat. Aerosp. Univ., Kharkov, Ukraine
Abstract :
This paper deals with one way to predict efficiency of despeckling for images acquired by a single-or multi-look synthetic aperture radar (SAR). A particular filter based on discrete cosine transform (DCT) and adapted to speckle characteristics is considered. Two quantitative criteria characterizing despeckling efficiency are analyzed. It is shown that even one parameter that can be calculated for image blocks sequentially or in parallel allows carrying out a rather accurate prediction. Moreover, such a prediction is possible for both spatially uncorrelated and correlated speckle.
Keywords :
discrete cosine transforms; image filtering; prediction theory; radar imaging; speckle; synthetic aperture radar; DCT-based filter; SAR image; despeckling efficiency prediction; discrete cosine transform; image block; quantitative criteria; speckle characteristic; synthetic aperture radar; Filtering; Fitting; Measurement; Noise; Noise reduction; Speckle; Standards; DCT; SAR imagery; denoising; speckle;
Conference_Titel :
Distributed Computing in Sensor Systems (DCOSS), 2015 International Conference on
Conference_Location :
Fortaleza
DOI :
10.1109/DCOSS.2015.16
