Comparative Study of Speckle Noise Reduction Approaches for Interferometric Synthetic Aperture Radar Images

Abstract

Speckle noise appearing in the interferometric SAR (InSAR) phase image degrades the quality of interferogram seriously and makes interferogram reflect the scattering characteristics of the target inaccurately, reducing the capability of extracting DEM information of target areas. Therefore, speckle noise reduction plays a major role in InSAR processing by using interferogram filtering. First, according to a terrain model with the assumed geometrical parameters in InSAR system, the paper simulated an interferometric SAR phase image with noise, which can be characterized by the multilook phase distribution based on the circular Gaussian assumption [1]. Second, the paper explores three interferogram filtering algorithm to remove speckle noise: Goldstain filter, Rotating Kernel Transformation and Lee filter; the proper implementation of three methods is given. Last, the paper discusses about the performance comparatively based on experimental results and gives broad conclusions and presents recommendations.