Skip to Main Content
An example based super-resolution algorithm for digital video using contourlet transform is proposed in this paper. The input is a low resolution video sequence together with a high resolution still image of similar content. Firstly, the low resolution frames are interpolated to the same spatial resolution as the reference still image. For the good properties of directional, multiscale and anisotropy, the nonsub-sampled contourlet is utilized to create the training set of transform coefficient patches from the high resolution still image. Block based motion estimation is then applied inside the complete training set to find the best matching between interpolated frame and reference still image. According to the correspondence between low frequency and high frequency pairs, the missing high frequency information of the input frame can be easily learned from the training set. Finally, an inverse contourlet transform is applied to the interpolated frame and supplement high frequency subbands to recover the super-resolved image. Preliminary experimental results on video frames show that the proposed super- resolution algorithm outperforms conventional spatial interpolation methods and wavelet based interpolation algorithm both in visual quality and the PSNR value.
Date of Conference: 22-24 Aug. 2007