JOINT FILTERING OF SAR INTERFEROMETRIC PHASE AND AMPLITUDE DATA IN URBAN AREAS BY TV MINIMIZATION

This paper investigates the use of a popular regularization model, the Total Variation minimization (TV), to lter SAR interferometric images (amplitude and phase data). This model has been extensively used for its property of edge preservation and is therefore well adapted for urban areas. Using a TV model adapted to multi-dimensionnal data, we propose to do a joint ltering of phase and amplitude images. Due to the many local minima, the minimization of such a model is hard to perform. A new fast approximate discrete algorithm is presented. Besides, the regularization strength is set automatically. The ltering can be applied as a preprocessing step for further image analysis steps such as 3D reconstruction. Results on real images are presented.