Abstract

In many digitized images with low resolution, the image is degraded more by aliasing than by the diffraction limit of the imaging system. Multiple low-resolution images, such as images of a moving object taken one after the other, can be combined to obtain a single image with much improved resolution by exploiting the fact that the information in each low-resolution image is slightly different from that of the others. The images are first centered on a common point with subpixel resolution by using correlation or absolute differences; then they are combined into a single image by means of a new technique that is superior to the improvement obtained by simply adding the images and is much faster than previous iterative methods. The new method involves interpolation from the low-resolution matrix of pixels to a higher-resolution matrix. Experiments have shown that the resolution gain is equivalent to that of previous methods, with a gain in speed of two orders of magnitude. The method is fast enough to be implemented on a personal computer.

© 1990 Optical Society of America