Research on Super Resolution Reconstruction Based on Scale Space

With the development of the electronics and information technology, the requirement of people toward the resolution of images become higher. However, the resolution of acquired images can´t meet our needs because of the different imaging system from a variety of factors. And it will cost too much to improve image resolution via using the physical hardware. So it requires a new method to solve this problem. Super-resolution is emerged in this context. This technique uses a software approach to enhance the resolution based on the existing images. Thus it has aroused widespread concern when it appeared. In this paper, on the basis of previous studies, a new super-resolution reconstruction method based on linear regression of scale-space which could obtain higher resolution image is proposed and the experiments indeed show the higher resolution image could be obtained after using the method.