Spline Surface In The Super-resolution Image Reconstruction

An important limitation of electronic imaging today is that most available still-frame or video cameras can only record images at a resolution lower than desirable. This is related to certain physical limitations of the image sensors, such as finite cell area and finite aperture time. Although higher-resolution imaging sensors are being advanced, these may be too expensive and/or unsuitable for mobile imaging applications.Super-resolution refers to obtaining video at a resolution higher than that of the camera (sensor) used in recording the image. Because most images contain sharp edges, they are not strictly band-limited. As a result, digital images usually suffer from aliasing due to undersampling, loss of high-frequency detail due to relative motion or out-of-focus. Supperresolution involves up-conversion of the input sampling lattice as well as reducing or eliminating aliasing and bluring. Superresolution from a single low-resolution, and possibly blurred, image is known to be highly ill-posed. However when a sequence of low-resolution frames is available, such as those obtained by a video camera, the problem becomes more manageable. It is evident that the 3-Dspatio-temporal sampling grid contains more information than any 2-D still-frame sampling grid. Interframe Superresolution methods exploit this additional information, contained in multiple frames, to reconstruct a high-resolution still image or a sequence o high-resolution images.The accurate subpixel motion information is indispensable to the process of Superresolution image reconstruction from a sequence of low resolution images. And it is a main task for computer vision and image processing to get the accurate subpixel motion information. We introduce bicubic spline surface fitting to the field of image registration. We use bicubic spline surface to simulate the local image block, and analysis the reason for edge blur. Then we propose the revised mathematics model, which enhances the image edge. On the base of precise description of digital image, we can use optimal method to get more accurate subpixel motion information. Finally, we use a iterative method to interpolated reconstruct the high resolution images.We apply our method in the Superresolution reconstruction process. The numeric results and the distinct reconstructed images prove the feasible of our method.