Self-sampled image resolution enhancement using dual-tree complex wavelet transform

Abstract

In this paper, a dual-tree complex wavelet transform domain image resolution enhancement method is proposed. The method estimates detail wavelet coefficients for the input low-resolution (LR) image using different types of deformations on the initial estimate of high-resolution (HR) image. Edge preserving smoothing filtering with different parameters is used in deformations. Decomposition of each deformed HR image results in a different set of detail wavelet coefficients for LR image, and the resultant HR image is computed by averaging the different reconstructions from LR image using different detail wavelet coefficient sets. The perceptual and objective quality of resolution enhanced images compare favorably with recently emerged methods in the field. © EURASIP, 2009.