A commonly pronounced artifact in many remote sensing (satellite) images are stripes, mainly due to the variations of calibrations, sensor arrangement and the view angles from detector to detector. We present a destriping method by developing a functional with anisotropic smoothing transverse to the stripes, while preserving the rest of the features of the image. The destriped image is uncovered by optimizing the functional using variational methods, followed by numerical methods from finite difference approximations. We demonstrate our method on two different striped data sets (VIIRS and HICO).
A commonly pronounced artifact in many remote sensing (satellite) images are stripes, mainly due to the variations of calibrations, sensor arrangement and the view angles from detector to detector. We present a destriping method by developing a functional with anisotropic smoothing transverse to the stripes, while preserving the rest of the features of the image. The destriped image is uncovered by optimizing the functional using variational methods, followed by numerical methods from finite difference approximations. We demonstrate our method on two different striped data sets (VIIRS and HICO).