Motion blur correction
First Claim
1. A method of correcting blur in a motion blurred image using an estimate of motion blur direction and motion blur extent based on the motion blurred image, the method comprising:
- generating an initial guess image based on said motion blurred image;
blurring the guess image as a function of said estimated blur direction and blur extent;
comparing the blurred guess image with the motion blurred image to generate an error image;
blurring and weighting the error image; and
combining the error image and the initial guess image thereby to update the guess image and correct for blur.
2 Assignments
0 Petitions
Accused Products
Abstract
A method and apparatus of correcting blur in a motion blurred image includes estimating the direction of blur in the motion blurred image based on edge response of the motion blurred image over a set of discrete directions and over subgroups of the discrete directions. The extent of blur in the motion blurred image is also estimated. An initial guess image based on the motion blurred image is generated and the guess image is blurred as a function of the estimated blur direction and blur extent. The blurred guess image is compared with the motion blurred image to generate an error image. The error image is blurred and weighted and then combined with the initial guess image thereby to update the guess image and correct for blur.
-
Citations
48 Claims
-
1. A method of correcting blur in a motion blurred image using an estimate of motion blur direction and motion blur extent based on the motion blurred image, the method comprising:
-
generating an initial guess image based on said motion blurred image;
blurring the guess image as a function of said estimated blur direction and blur extent;
comparing the blurred guess image with the motion blurred image to generate an error image;
blurring and weighting the error image; and
combining the error image and the initial guess image thereby to update the guess image and correct for blur. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13)
-
-
14. A method of correcting blur in a motion blurred image comprising:
-
estimating the direction of blur in said motion blurred image based on edge response of said motion blurred image over a set of discrete directions extending through said motion blurred image and over subgroups of said discrete directions;
estimating the extent of blur in said motion blurred image;
generating an initial guess image based on said motion blurred image;
blurring the guess image as a function of said estimated blur direction and blur extent;
comparing the blurred guess image with the motion blurred image to generate an error image;
blurring the error image; and
combining the error image and the initial guess image thereby to update the guess image and correct for blur. - View Dependent Claims (15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35)
-
-
36. A method of estimating blur direction in a motion blurred image comprising:
-
determining the deviation between the edge response of said motion blurred image over a set of discrete spaced directions extending through said motion blurred image and the edge response of said motion blurred image over possible subgroup combinations of the discrete directions in said set, each subgroup combination including a pair of quadrants with one quadrant comprising a plurality of consecutive discrete directions;
for the subgroup combination yielding the highest deviation declaring the direction that bisects the one quadrant as the blur direction when the edge response over the discrete directions in the one quadrant is less than the edge response over the discrete directions in the other quadrant; and
otherwise declaring the direction normal to the direction that bisects the one quadrant as the blur direction. - View Dependent Claims (37, 38, 39, 40)
-
-
41. An apparatus for estimating blur direction in a motion blurred image comprising:
-
an edge detector determining the edge response through said motion blurred image along a plurality of different discrete directions forming a set;
a deviation determinor determining the deviation between the edge response of said motion blurred image over said set of discrete spaced directions and the edge response of said motion blurred image over possible subgroup combinations of said set, each subgroup combination including a pair of quadrants with one quadrant comprising a plurality of consecutive discrete directions; and
a comparator determining the subgroup combination yielding the highest deviation, said comparator for that determined subgroup combination declaring the direction that bisects the one quadrant as the blur direction when the edge response over the discrete directions in the one quadrant is less than the edge response over the discrete directions in the other quadrant and otherwise declaring the direction normal to the direction that bisects the one quadrant as the blur direction. - View Dependent Claims (42, 43, 44, 45)
-
-
46. An apparatus for correcting blur in a motion blurred image using an estimate of motion blur direction and motion blur extent based on the motion blurred image, said apparatus comprising:
-
a motion blur filter blurring an initial guess image based on said motion blurred image as a function of said estimated blur direction and blur extent;
a comparator comparing the blurred guess image with the motion blurred image and generating an error image, said motion blur filter further blurring the error image and weighting the error image according to an error weighting factor; and
an adder combining the error image and the initial guess image thereby to update the guess image and correct for blur. - View Dependent Claims (47, 48)
-
Specification