Image Super-resolution Method Using Registration of Multi-scale Wavelet Components with Consideration of Digital Cinema Noise
Yasutaka Matsuo, Ryoki Takada, Shinya Iwasaki, Jiro Katto · The Journal of The Institute of Image Information and Television Engineers · 2014
Image super-resolution quality is degraded by the noise component of an image. However the noise should not be eliminated because of its importance in high-definition impression for digital cinema. We therefore propose an image super-resolution method using the synthesis of the super-resolution signal and noise components respectively after dividing an original image into signal and noise components. The signal component's resolution is enhanced using registration between the signal component and its multi-scale Wavelet components with resolution-enhanced parameter optimization.