Method for multiframe Wiener restoration of noisy and blurred image sequences

Claim 2

A cross-correlated multiframe (CCMF) Wiener filtering method for restoring image sequences comprising: a) digitizing a sequence of images; b) dividing the digitized sequence of images into groups of N frames; c) estimating power spectra and cross power spectra of the N digitized image frames; d) providing as inputs to a cross-correlated multiframe Wiener filter: the estimation of the power spectra and cross power spectra of the N digitized frames, noise power spectra of the N digitized frames, and blur parameters; e) performing cross-correlated multiframe Wiener filtering using said inputs to obtain N restored image frames as an output; f) repeating steps b-e for every other group of N frames in the digitized sequence of images to produce a restored image sequence; and g) providing the digitized restored image sequence to an output device.

Claim 3

The method of Claim 1, wherein the noise power spectra of the N digitized images is known a priori.

Claim 4

The method of Claim 1, wherein the noise power spectra of the N digitized images is calculated from the N digitized images themselves by computing sample variances over uniform regions of the image frames.

Claim 5

The method of Claim 1, wherein the power spectra and cross power spectra of the N digitized image frames in a given group are estimated from a set of sharp prototype image frames.

Claim 6

A cross-correlated multiframe (CCMF) Wiener filtering method for restoring image sequences comprising: a) digitizing a sequence of images; b) dividing the digitized sequence of images into groups of N frames; c) independently restoring the N digitized frames using single-frame Wiener restoration; d) calculating power spectra and cross power spectra of the N digitized frames of the sequence using the image sequence formed by the resulting image frames as a prototype image sequence; e) providing as inputs to a cross-correlated multiframe Wiener filter: the estimation of the power spectra and cross power spectra of the N digitized frames, noise power spectra of the N digitized frames, and blur parameters; f) performing cross-correlated multiframe Wiener filtering using said inputs to obtain N restored image frames as an output; g) repeating steps b-f for every other group of N frames in the digitized sequence of images to produce a restored image sequence; and h) providing the digitized restored image sequence to an output device.

Claim 7

A cross-correlated multiframe (CCMF) Wiener filtering method for restoring image sequences comprising: a) digitizing a sequence of images; b) dividing the digitized sequence of images into groups of N frames; c) independently restoring the N digitized frames using single-frame Wiener restoration; d) calculating power spectra and cross power spectra of the N digitized frames of the sequence using the image sequence formed by the resulting image frames as a prototype image sequence; e) providing as inputs to a cross-correlated multiframe Wiener filter: the estimation of the power spectra and cross power spectra of the N digitized frames, noise power spectra of the N digitized frames, and blur parameters; f) performing cross-correlated multiframe Wiener filtering using said inputs to obtain N restored image frames as an output; g) repeating steps b-f for every other group of N frames in the digitized sequence of images to produce a restored image sequence; and h) providing the digitized restored image sequence to an output device.

Claim 8

The method of Claim 5, wherein the noise power spectra of the N digitized images is known a priori.

Claim 9

The method of Claim 5, wherein the noise power spectra of the N digitized images is calculated from the N digitized images themselves by computing sample variances over uniform regions of the image frames.

Claim 10

The method of Claim 1 or 5, wherein the blur parameters are known a priori.

Claim 11

The method of Claim 1 or 5, wherein the blur parameters are estimated by the log-spectrum method.

Claim 12

The method of Claim 1 or 5, wherein the blur parameters are estimated by the cepstrum method.

Claim 13

The method of Claim 1 or 5, wherein the step e) or f) includes the step of computing discrete Fourier transform samples of estimates of the N digitized image frames in parallel and computing inverse discrete Fourier transform to obtain image frame estimates in a spatial domain.

Claim 14

The method of Claim 11, wherein the step of computing discrete Fourier transform samples of estimates of the N digitized image frames in parallel is performed according to the equation: where F i,k are elements of two-dimensional discrete Fourier transform of the estimate of the i th frame, i = 1,2,···,N, P f;ip,k are the elements of the diagonalized cross-correlation matrix of the actual i th and the p th frames of the N digitized frames, are the elements of the diagonalized blur matrix for the p lh frame, and Zpq, k are the elements of an M 2 x M 2 matrix that are determined by inverting an N x N matrix, in general, whose elements are defined in terms of the elements of blur, diagonalized cross-correlation, and diagonalized noise autocorrelation matrices, and Gq, k are the elements of two-dimensional discrete Fourier transform of the degraded version of a q lh frame.

Claim 15

The method of Claim 11, wherein step e) includes the step of computing discrete Fourier transform samples of estimates of the N digitized image frames in parallel is performed according to the equation: where F ,k are elements of two-dimensional discrete Fourier transform of the estimate of the i th frame, i = 1,2,···,N, P f;ip,k are the elements of the diagonalized cross-correlation matrix of the actual i th and the p th frames of the N digitized frames, are the elements of the diagonalized blur matrix for the p th frame, and Zpq, k are the elements of an M 2 x M 2 matrix that are determined analytically in terms of the elements of blur, diagonalized cross-correlation, and diagonalized noise autocorrelation matrices, and Gq, k are the elements of two-dimensional discrete Fourier transform of the degraded version of a q lh frame.

Claim 16

A motion-compensated multiframe (MCMF) Wiener filtering method for restoring image sequences comprising; a) digitizing a sequence of images; b) dividing the digitized sequence of images into groups of N neighboring frames for every flame in the sequence of images; c) selecting a frame to be restored from a group of N frames to be used as a reference frame; d) estimating a power spectrum of the actual version of the reference frame using a sharp image other than the actual reference frame as a prototype image; e) estimating a noise power spectrum of the reference frame; f) using a motion estimation algorithm to estimate motion vectors between the reference frame and the other frames in the group of N frames, said motion vectors carrying interframe correlation information; g) providing as inputs to a motion-compensated multiframe Wiener filter: the estimation of the power spectrum of the actual version of the reference frame, the estimation of motion vectors, noise power spectrum of the reference frame, and blur parameters; h) performing a motion compensated multiframe Wiener filtering using said inputs to obtain a restored reference frame as an output; i) repeating steps b-h for every other frame in the sequence to restore the entire image sequence; and j) providing the restored digitized image sequence to an output device.

Claim 17

A motion-compensated multiframe (MCMF) Wiener filtering method for restoring image sequences comprising; a) digitizing a sequence of images; b) dividing the digitized sequence of images into groups of N neighboring frames for every flame in the sequence of images; c) selecting a frame to be restored from a group of N frames to be used as a reference frame; d) estimating a power spectrum of the actual version of the reference frame using a sharp image other than the actual reference frame as a prototype image; e) estimating a noise power spectrum of the reference frame; f) using a motion estimation algorithm to estimate motion vectors between the reference frame and the other frames in the group of N frames, said motion vectors carrying interframe correlation information; g) providing as inputs to a motion-compensated multiframe Wiener filter: the estimation of the power spectrum of the actual version of the reference frame, the estimation of motion vectors, noise power spectrum of the reference frame, and blur parameters; h) performing a motion compensated multiframe Wiener filtering using said inputs to obtain a restored reference frame as an output; i) repeating steps b-h for every other frame in the sequence to restore the entire image sequence; and j) providing the restored digitized image sequence to an output device.

Claim 18

The method of claim 14, wherein the motion estimation algorithm is the Fogel algorithm.

Claim 19

A cross-correlated multiframe (CCMF) Wiener filtering method for restoring image sequences comprising a) digitizing a sequence of images; b) dividing the digitized sequence of images into groups of N frames; c) estimating power spectra and cross power spectra of the N digitzed image frames; d) providing as inputs to a cross-correlated multiframe Wiener filter, e) performing cross-correlated multiframe Wiener filtering using said inputs to obtain N restored image frames as an output; and f) providing the digitized restored image sequence to an output device.

Claim 20

A cross-correlated multiframe (CCMF) Wiener filtering method for restoring image sequences comprising a) digitizing a sequence of images; b) dividing the digitized sequence of images into groups of N frames; c) estimating power spectra and cross power spectra of the N digitzed image frames; d) providing as inputs to a cross-correlated multiframe Wiener filter, e) performing cross-correlated multiframe Wiener filtering using said inputs to obtain N restored image frames as an output; and f) providing the digitized restored image sequence to an output device.