Super-resolution mapping using Hopfield Neural Network with Panchromatic image

Super-resolution mapping or sub-pixel mapping is a set of techniques to produce the hard land cover map at sub-pixel spatial resolution from the land cover proportion images obtained by soft-classification methods. In addition to the information from the land cover proportion images at the original spatial resolution, supplementary information at the higher spatial resolution can be used to produce more detailed and accurate sub-pixel land cover maps. In this research, panchromatic (PAN) imagery was used as an additional source of information for super-resolution mapping using the Hopfield neural network (HNN). Forward and inverse models based on the local end-member spectra were incorporated in the HNN to support a new panchromatic reflectance constraint added to the energy function. A set of simulated data were used to test the new technique. The results suggest that 1 m IKONOS panchromatic imagery can be used as supplementary data to increase the detail and accuracy of the sub-pixel land cover maps produced by superresolution mapping of a 4 m land cover proportion image.