Implementation of image super resolution based on POCS method
no vote
<span style="color:#4F4F4F;font- family:-apple-system , "font- size:16px;background-color :#FFFFFF;"> The low resolution image is filtered by Gaussian filter (to eliminate the influence of noise), and then the integer displacement is estimated on the filtered image (which is equivalent to using large image block to estimate large displacement), and the estimated displacement is used as the initial value of the next level matching. Next, the low resolution image is sampled by bilinear interpolation, and the up sampled image is filtered by Gaussian filter (to eliminate the data instability caused by bilinear interpolation), and then the block matching is continued on the filtered image to obtain the sub-pixel precision motion vector. In this way, through step-by-step up sampling, Gaussian filtering and block matching, the required matching accuracy can be obtained.