An interactive super resolution coding approach of images and videos for enhanced user visualization

In this paper we propose an interactive super resolution coding technique to enhance user's visualization of received low resolution (LR) images. The proposed technique hides/embeds high frequencies and edges in the LR image, and constructs a high resolution (HR) image upon a user click. The proposed technique is suitable for many applications (eg. digital libraries) that display images and videos to the user in small sizes, and should be capable of reconstructing a HR version of the image or video upon a click on the LR version. The proposed approach consists of an insertion part and an extraction part. In the insertion part (that could be treated as an acquisition algorithm), high frequency components of the sensed image are embedded/inserted among the content of the finally delivered LR image in an invisible manner. The resulting sensed image is a regular low resolution image, however when performing the extraction part on it, a HR image is constructed without any side information in a blind, fast, robust and effective manner. The proposed approach has been tested on several images and video sequences against noise and compression, and gave very competitive results with current literature methods.