Method for producing enhanced-resolution image by use of a plurality of low-resolution images

A method for producing an enhanced-resolution image by use of M first low-resolution images associated with a scene. M being a natural number larger than 1, each of the M first images being represented by a set of uniformly sampled and quantized pixels, a relationship of a magnification factor existing between the enhanced-resolution image and the M first images, said method comprising the steps of:

• (a) storing the M first images;
• (b) selecting one from the M first images as a prototype image, and
• referring the non-selected (M-1) first images to being as (M-1) second images;
• (c) based on the magnification factor, interpolating extra pixels into the set of pixels of the prototype image, and then inferring the values of the extra pixels according to the values of neighbor pixels thereof to produce an interpolated prototype image, and calculating a respective translation between each of the (M-1) second images and the interpolated prototype image;
• (d) dividing the translations of the (M-1) second images by the magnification factor, respectively, to obtain the modulus with respect to the translations of the (M-1) second images, selecting, based on a criteria, one from the second images whose related modulus are the same, and referring the selected second images together with the rest of the second images whose related modulus are not the same to being as N third images, wherein N is equal to or less than (M-1);
• (e) down-sampling the interpolated prototype image N times according to the respective translation between each of the N third images and the interpolated prototype image to produce N fourth images which each corresponds to one of the N third images;
• (f) calculating the difference between each of the N third images and the corresponding fourth image thereof;
• (g) adjusting the values of the pixels of the interpolated prototype image according to an average of the differences calculated in step (f);
• and
• (h) repeating step (e) through step (g), until the values of the pixels of the interpolated prototype image converge to a satisfactory result, and
• referring the interpolated prototype image whose values of pixels converge to the satisfactory result to being as the enhanced-resolution image;
• wherein step (e) is performed according to the following formula:
• gk(n)=(Tk(ƒ(n))·h)↓s,
• wherein gk(n) is the kth forth image at the nth down-sampling process, ƒ(n) is the interpolated prototype image been adjusted n times, h is the blurring operator defined by a point-spread-function, Tk is the transformation operator with respect to the translation between the kth third image and the interpolated prototype image, ↓s is the down-sampling operator.