Spectral Interpolation on 3 × 3 Stencils for Prediction and Compression

摘要: Many scientific, imaging, and geospatial applications produce large high-precision scalar fields sampled on a regular grid. Lossless compression of such data is commonly done using predictive coding, in which weighted combinations previously coded samples known to both encoder decoder are used predict subsequent nearby samples. In hierarchical, incremental, or selective transmission, the spatial pattern neighbors often irregular varies from one sample next, precludes prediction based single stencil fixed set weights. To handle situations make best use available neighboring samples, we propose local spectral predictor that offers optimal by tailoring weights each configuration These may be precomputed stored small lookup table. We show through several coding our improves for various sources data.