A novel object based Super-Resolution using Phase Template Matching

In many applications, a high-resolution image is desired from available multiple degraded image frames, especially in surveillance recording. This paper represents a new object based Super-Resolution using Phase Template Matching. This is a problem of extracting an object from a series of images which extracted image has more high frequency information than any one of the input images. The desired object will be detected in every images of sequence to produce a sequence of images of desired object. Then these images are put together by means of super-resolution technique that makes the image of that object with more high frequency details as well as more pixels. The super-resolution technique is based on Generalized Majorization Minimization approach.