Enhancing Spatial Resolution of Hyperspectral Imagery Using Sensor's Intrinsic Keystone Distortion

摘要

In this paper, we develop a novel technology that can enhance the spatial resolution of hyperspectral data cube without using any additional images, as would be the case in image fusion. The technology exploits interband spatial misregistration or distortion (often referred to as “keystone”) of the sensor that acquired the data cube and uses it as additional information to increase the spatial resolution of the data cube. Three methods have been developed to derive subpixel-shifted images from the data cube itself in exploiting the sensor's intrinsic characteristics. Two schemes were proposed to organize the derived subpixel-shifted images before being integrated into the high-resolution image using the iterative-back-projection fusion. Experimental results show that the technology can enhance the spatial resolution in the cross-track direction of hyperspectral data cubes by a factor of two.