Super-Resolution Image Reconstruction Based on Improved POCS Algorithm
Projections onto convex sets (POCS) algorithm is a widely used super-resolution image reconstruction method. Aiming at the edge ringing effect of traditional POCS algorithm, this paper analyzes the basic reason causing the effect, and adopts an improved POCS algorithm to reduce it. In the improved algorithm, the Point Spread Function (PSF) centered at any edge pixel is weighted, making the far the position of the PSF coefficient is from the edge, the smaller the corresponding PSF coefficient is, and the coefficients remain unchanged along the edge direction. This paper uses wavelet transform modulus maxima method to detect image edges. Considering the edge ringing effect is not only relevant to the edge pixels, but also relevant to their neighboring pixels, we dilate the edge-detected image with a structuring element so as to obtain thicker image edges. Experimental results show that our method greatly reduces the edge ringing effect at little cost in terms of image sharpness, so we can get a better reconstruction image.
KeywordsSuper-resolution image reconstruction Projections onto convex sets Wavelet transform modulus maxima Edge ringing
- 1.Youla DC, Webb H (1982) Image restoration by the method of convex projections: part 1-theory. IEEE Trans Med Imag Mi- 1(2):81–94Google Scholar
- 2.Sezan MI, Stark H (1982) Image restoration by the method of convex projections: part 2-applications and numerical results. IEEE Trans Med Imag Mi 1(2):95–101Google Scholar
- 4.Tekalp M, Ozkan MK, Sezan MI (1992) High-resolution image reconstruction from lower-resolution image sequences and space-varying image restoration. IEEE Int Conf Acoustics Speech Signal Processing 3:169–172Google Scholar