Description:

Computer Vision is the ability of computers to understand and analyze imagery content in the same way humans subjectively do.  In satellite meteorology, there are a myriad of computer vision algorithms which are used to objectively quantify properties of the atmosphere and oceans that are remotely sensed, such as fluid motions, cloud height distributions, and cloud/snow/ice/shadow discrimination, significantly increasing the benefits of standalone satellite imagery.  This session topic solicits presentations which cover development of computer vision algorithms and applications for satellite meteorology, such as optical (or scene) flow, image stereoscopy, novel view generation, temporal interpolation, scene (or pixel) classification, segmentation, object/feature detection, image compositing, and 3D reconstruction.  We highly encourage presentations that demonstrate the value of such algorithms for diagnosis, nowcasting, and forecasting of meteorological and oceanic phenomena.