Change detection of water-body using Chinese GF-3 polarimetric SAR images

Abstract

Change detection and monitoring of water areas using synthetic aperture radar (SAR) datasets is becoming an important research topic for SAR Earth observation. And polarimetric SAR images are more advantageous when detecting water than are single-polarization SAR images. To accomplish water change detection based on polarimetric SAR data, a multi-class water change detection method based on the complex Wishart distribution is proposed in this paper. Considering that the SPAN image can efficiently suppress noise and portray meticulous image structure, and that the water SPAN value is much smaller than others in the image, as well as the fact that the covariance matrix of polarimetric SAR image obeys the complex Wishart distribution, this paper takes the negative logarithm of the PDF of every pixel to reconstruct two new images. Based on the characteristics of the complex Wishart distribution, a joint map under certain conditions is constructed. Then, the ratio image between one of the two new images and the joint image contains three change areas: an unchanged area, one in which the dark domain becomes the bright domain, and the inverse: one in which the bright domain becomes the dark domain. Using these four images, the water level change can be divided into six categories that contain the variations for semi-flooded areas. Two scenes of Chinese Gaofen-3 (GF-3) polarimetric SAR data are used to validate our method. The experimental results show that the proposed method can detect the water changes successfully.