Comparative Analysis of Edge Based Single Image Superresolution

摘要

Superresolution image reconstruction provides an effective way to increase image resolution from a single or multiple low resolution images. There exists various single image superresolution based on different assumptions, amongst which edge adaptive algorithms are particularly used to improve the accuracy of the interpolation characterizing the edge features in a larger region. A recent algorithm for image iterative curvature based interpolation (ICBI) performs iterative procedure of the interpolated pixels obtained by the 2nd order directional derivative of the image intensity. ICBI in comparison with bicubic interpolation and the other interpolation algorithm such as improved new edge directed interpolation (INEDI) provides notably higher values in terms of qualitative and quantitative analysis. Comparative analysis of these algorithms performed on number of test images on the basis of PSNR and RMSE metrics show effectiveness of edge based techniques.