Object-Based Area-to-Point Regression Kriging for Pansharpening

Optical earth observation satellite sensors often provide a coarse spatial resolution (CR) multispectral (MS) image together with fine (FR) panchromatic (PAN) image. Pansharpening is technique applied to such sensor images generate an FR MS by injecting detail taken from the PAN while simultaneously preserving spectral information of methods are mostly on per-pixel basis and use extract detail. However, many land cover objects in not illustrated as independent pixels, but spatially aggregated pixels that contain important semantic information. In this article, object-based pansharpening approach, termed area-to-point regression kriging (OATPRK), proposed. OATPRK aims fuse at scale and, thus, takes advantage both unified within CR composed three stages: segmentation, regression, residual downscaling. Three data sets acquired IKONOS Worldview-2 11 benchmark algorithms were used comprehensive assessment proposed approach. synthetic real experiments, produced most superior pan-sharpened results terms visual quantitative assessment. new conceptual method advances pixel-level geostatistical approach object level provides more accurate images.