Morphology based iterative back-projection for super-resolution reconstruction of image

Abstract

Super-resolution (SR) reconstruction using iterative back projection (IBP) is a well-known and computationally efficient method for the enhancement of spatial resolution of an image. However, IBP algorithm has some limits in the performance like ringing artifacts in the strong edge area of an image. In this paper, we propose an improved algorithm that modify IBP based SR reconstruction method enable more detail reconstruction and to lessen the ringing artifacts in the image. The current task manages with a constrained optimization of the SR reconstruction problem enforcing the provincially adaptive edge regularization technique using mathematical morphology in the iterative process. Adding to this, cuckoo search and gradient search algorithm combining a hybrid optimization is used to minimize the overall reconstruction error from the high resolution solution of previous IBP model. Experimental results reveal the effectiveness of proposed algorithm it's not only reducing the ringing artifacts, but it is also preserving the edges for getting better resolution and visual perception as compared to the existing state of art methods. It is also clear that this hybrid optimization technique doing something to a greater degree to the corresponding individual search methods.