Graph-Based Interpolation for Remote Sensing Data

Remote sensing technologies, such as radiometry radar or synthetic-aperture radar (SAR) provide satellite observations with different spatio-temporal resolutions. Some of these observations can be combined to enhance the resolution of satellite products. In this paper, we formulate the data interpolation problem as a signal reconstruction on a graph, where the weights of edges between nodes are chosen as a function of high resolution ancillary data, e.g., geospatial data or observations from an instrument providing high resolution imagery. Our method is initialized with a high resolution signal directly interpolated from coarse observations obtained from a first instrument. Then, we solve a quadratic Laplacian optimization problem to ensure that the obtained high resolution observations are consistent with our original estimates while being smooth on the constructed graph. Preliminary results show that our approach leads to better remote sensed data reconstruction when compared against traditional spatial filtering and interpolation techniques.