Method and apparatus to correct digital image blur due to motion of subject or imaging device

• US 8,064,720 B2
• Filed: 11/19/2008
• Issued: 11/22/2011
• Est. Priority Date: 03/25/2004
• Status: Active Grant

• Current Assignee: Clear Imaging Research, LLC
• Original Assignee: Fatih M. Ozluturk
• Inventors: Ozluturk, Fatih M.
• Primary Examiner(s): Bhatnagar, Anand
• Assistant Examiner(s): Tsai, Tsung-Yin
• Application Number:
  US12/274,096
• Publication Number:
  US 20090128641A1
• Time in Patent Office: 1,098 Days
• Field of Search: None
• US Class Current: 382/275
• CPC Class Codes: G02B 27/646 compensating for small deviations, e.g. due to vibration or shake movement of one or more optical elements for control of motion blur in cameras, projectors or printers G03B2205/0007; image stabilisation in cameras peculiar to the presence or use of an electronic image sensor H04N23/68
• CPC Class Codes: G06T 2207/10004 Still image; Photographic image
• CPC Class Codes: G06T 2207/10016 Video; Image sequence
• CPC Class Codes: G06T 2207/20024 Filtering details
• CPC Class Codes: G06T 2207/20201 Motion blur correction
• CPC Class Codes: G06T 2207/20212 Image combination
• CPC Class Codes: G06T 5/70 Denoising; Smoothing
• CPC Class Codes: G06T 5/73 Deblurring; Sharpening
• CPC Class Codes: G06T 7/20 Analysis of motion motion estimation for coding, decoding, compressing or decompressing digital video signals H04N19/43, H04N19/51
• CPC Class Codes: H04N 23/62 Control of parameters via user interfaces
• CPC Class Codes: H04N 23/68 for stable pick-up of the scene, e.g. compensating for camera body vibrations
• CPC Class Codes: H04N 23/681 Motion detection
• CPC Class Codes: H04N 23/6811 based on the image signal
• CPC Class Codes: H04N 23/6812 based on additional sensors, e.g. acceleration sensors
• CPC Class Codes: H04N 23/683 performed by a processor, e.g. controlling the readout of an image memory
• CPC Class Codes: H04N 23/687 by shifting the lens or sensor position
• CPC Class Codes: H04N 23/80 Camera processing pipelines; Components thereof
• CPC Class Codes: H04N 25/615 involving a transfer function modelling the optical system, e.g. optical transfer function [OTF], phase transfer function [PhTF] or modulation transfer function [MTF]
• CPC Class Codes: H04N 5/265 Mixing

Abstract

Signal processing techniques are applied to digital image data to remove the distortion caused by motion of the camera, or the movement of the subject being photographed, or defective optics, or optical distortion from other sources. When the image is captured, the effect of relative motion between the camera and the subject is that it transforms the true image into a blurred image according to a 2-dimensional transfer function. The 2-dimensional transfer function representing the motion is derived using blind estimation techniques or by using information from sensors that detect the motion. The transfer function is inverted and used to define a corrective filter. The filter is applied to the image and the blur due to the motion is removed, restoring the correct image. Another embodiment uses the transfer function to avoid blur by combining multiple consecutive images taken at a fast shutter speed.

Claims

1. A method, comprising:

a recording medium recording image data that represents an image;

a processor determining an initial state for an inverse transfer function;

the processor updating the inverse transfer function using an iterative technique, wherein the iterative technique includes, for each of a plurality of iterations;

determining a first estimate of the image data based on the inverse transfer function and a finite impulse response (FIR) filter;

determining a second estimate of the image data by applying a non-linear estimation function to the first estimate of the image data;

determining an estimation error parameter by comparing the second estimate of the image data and the image data; and

updating the inverse transfer function based on the estimation error parameter and a step-size parameter;

the processor determining a deconvolution filter based on the inverse transfer function; and

the processor filtering the image data using the deconvolution filter.

2. The method of claim 1, wherein the iterative technique further includes, for one or more of the plurality of iterations:

determining whether to perform a next iteration based on the estimation error parameter.

3. The method of claim 1, wherein the inverse transfer function includes one or more coefficients, and wherein the updating the inverse transfer function includes updating the one or more coefficients.

4. The method of claim 1, wherein the second estimate of the image data is a Bayes estimate.

5. The method of claim 1, wherein the transfer function represents motion in a portion of the image data that is smaller than the entire image data, and wherein the processor filtering the image data using the deconvolution filter is performed on the entire image data.

6. The method of claim 1, wherein the method is performed in a digital camera.

7. An apparatus, comprising:

a recording medium configured to record image data that represents an image;

a processor configured;

to determine an initial state for an inverse transfer function;

to update the inverse transfer function using an iterative technique, wherein the iterative technique includes, for each of a plurality of iterations;

determining a first estimate of the image data based on the inverse transfer function and a finite impulse response (FIR) filter;

determining a second estimate of the image data by applying a non-linear estimation function to the first estimate of the image data;

determining an estimation error parameter by comparing the second estimate of the image data and the image data; and

updating the inverse transfer function based on the estimation error parameter and a step-size parameter;

to determine a deconvolution filter based on the inverse transfer function; and

to filter the image data using the deconvolution filter.

8. The apparatus of claim 7, wherein the iterative technique further includes, for one or more of the plurality of iterations:

determining whether to perform a next iteration based on the estimation error parameter.

9. The apparatus of claim 7, wherein the inverse transfer function includes one or more coefficients, and wherein the updating the inverse transfer function includes updating the one or more coefficients.

10. The apparatus of claim 7, wherein the transfer function represents motion in a portion of the image data that is smaller than the entire image data, and wherein the processor filtering the image data using the deconvolution filter is performed on the entire image data.

11. The apparatus of claim 7, wherein the second estimate of the image data is a Bayes estimate.

12. The apparatus of claim 7, wherein the apparatus is a digital camera.