Document Type : علمی - پژوهشی

Authors

• Hamid Jannati ¹
• Mohammad Javad Valadan Zoej ²

¹ Department of Photogrammetry and Remote sensing, College of Geodesy and Geomatics, K. N. Toosi University of Technology

² Professor in Department of Photogrammetry and Remote sensing, College of Geodesy and Geomatics, K. N. Toosi University of Technology

Abstract

Speckle in Synthetic aperture radar images makes grainy effects, because of the coherent imaging system which cause some difficulties in object-oriented processes, like segmentation or classification. Therefore, a lot of methods have been developed for speckle reduction purpose. These methods can be classified but not limited in some approaches, like spatial based, transform based and optimization, which mostly suffer from limitations like edge and texture destruction and also regulating parameter dependence. In this paper a new structure has been presented based on adaptive filtering of the amplitude response of the discrete fourier transform of the image in the frequency space, which not only reduces the speckle but also preserves edges and delicate textures. In addition, it has low level of computation and complexity compared to the kernel dependent spatial approaches. The main contribution of the paper is to fit a predefined analytical function to amplitude response of the discrete fourier transform of the image, in order to recover underlying speckle reduced SAR image. Proposed method, improves equivalent number of looks index 50 percent and edge preservation index 50 and 30 percent for real and simulated synthetic aperture radar images, respectively.

Keywords

• SAR images
• speckle reduction
• frequency domain
• Fourier transform
• amplitude spectrum
• non-linear filtering