Image super resolution using interpolation and edge extraction in wavelet transform space
Herminio Chavez-Roman, Volodymyr I. Ponomaryov, Ricardo Peralta-Fabi · 2012
A novel approach in image enhancement is proposed. In contrast to other state-of-art algorithms, the designed technique is based on additional edges preservation procedure and mutual interpolation between the high frequency (HF) sub-band image performed via the discrete wavelet transform (DWT) and the input low resolution (LR) image. Finally, all sub-band images are combined to generate a new high-resolution (HR) image via Inverse DWT (IDWT). Numerous simulation results confirm that novel super resolution framework performs well in terms of objective criteria (PSNR, MAE and SSIM) and the more subjective measure of human vision in the different images, improving resolution sufficiently better than do existing methods.