Generalized Interpolation for Super-Resolution
Deepu Rajan, Subhasis Chaudhuri · Kluwer Academic Publishers eBooks · 2005
In this chapter, we present a generalized interpolation scheme for image expansion and generation of super-resolution images. The underlying idea is to decompose the image into appropriate subspaces, to interpolate each of the subspacs individually and finally, to transform the interpolated values back to the image domain. This method is shown to presere various optical and structural properties of the image, such as 3-D shape of an object, regional homogeneity, local variations in scene reflectivity, etc. The motivation for doing so has also been explained theoretically. The generalized interpolation scheme is also shown to be useful in perceptually based high resolution representation of images where interpolation is done on individual groups as per the perceptual necessity. Further, this scheme is also applied to generation of highresolution transparencies from low resolution transparencies.