Performance Evaluation Of Super-Resolution In Image Sequences

The technology of super resolution in image sequences uses superfluous information among image sequences to rebuild an image with higher resolution. We can get higher resolution from two ways, one is improving hardware, the other is processing image in software. Amelioration of hardware is costly, so super resolution is observing after its appearance. Today, super-resolution has been used in much fields.Super-resolution processing in image sequences is comprised of three aspects. The first is super-resolution algorithm; the second is precision focus in digital camera to prove quality of image before processing of image sequences; the third is performance evaluation of super-resolution. The paper studies these aspects. In aspect of precision focus in digital camera, to gets best low-resolution image sequences, the paper gives a means to judge if the camera is in focus by statistic of fringe. In aspect of super-resolution processing, the paper analyses model algorithms and ascertains algorithms our use. In aspect of performance evaluation of super-resolution, studies the ways that performance evaluation of super-resolution and the least number of images (the sufficient of number of low-images) to get the best result. The experimental result of processing in some fact image sequences show it approvable.