Complex-valued region-based-coupling image segmentation neural networks and their applications to interferometric synthetic aperture radar image processing

摘要

Complex-valued region-based-coupling image segmentation neural networks are proposed for interferometric synthetic aperture radar (InSAR) image segmentatinon. They deal with the amplitude and phase information of InSAR systems as a combined complex-amplitude image. Thereby, not only the reflectance but also the distance or optical length are consistently taken into account for the segmentation. Segmentation experiments are successfully demonstrated. They are applicable as preprocessing modules to future radar systems for image acquisition in, for instance, invisible fire smoke places and intelligent transportation systems by generating a processed image that is more recognizable by human and recognition systems.