Super resolution mapping of satellite images using Hopfield neural networks

Abstract

Super resolution mapping is a set of techniques to increase the spatial resolution of a land cover map obtained by soft classification methods. Linear spectral unmixing have been developed to estimate the class composition of image pixels, but their output provides no indication of how these classes are distributed spatially within the instantaneous field of view represented by the pixel. The use of a Hopfield neural network to map the spatial distribution of classes more reliably using prior information of pixel composition determined from linear spectral unmixing was investigated. The output from the linear spectral unmixing which is a set of area proportion images for each land cover class is given as input to the HNN. The network converges to a minimum of the energy function which is defined by the goals and constraints of the super resolution mapping task. The minimum of the energy of the network represents the best guess map of the given satellite image. The technique was applied to both real and simulated Landsat images, and the resultant maps provided an accurate and improved representation of the area under study. The Hopfield neural network represents a simple, robust, and efficient technique, and results suggest that it is a useful tool for identifying land cover targets from remotely sensed imagery at the subpixel scale.