Method and computer program for detecting rotation and magnification of images

• US 6,711,303 B1
• Filed: 12/01/1999
• Issued: 03/23/2004
• Est. Priority Date: 12/01/1999
• Status: Expired due to Fees

• Current Assignee: Eastman Kodak Company
• Original Assignee: Eastman Kodak Company
• Inventors: Honsinger, Chris W.
• Primary Examiner(s): Wu, Jingge
• Assistant Examiner(s): Hesseltine, Ryan J.
• Application Number:
  US09/452,415
• Time in Patent Office: 1,574 Days
• Field of Search: G06/K.964, 382/276, 382/286, 382/289, 382/278, 382/297, 382/298, 382/264, 382/205, 358/3.28
• US Class Current: 382/289
• CPC Class Codes: G06T 1/0064 Geometric transfor invariant watermarking, e.g. affine transform invariant
• CPC Class Codes: G06T 2201/0051 Embedding of the watermark in the spatial domain
• CPC Class Codes: G06T 2201/0065 Extraction of an embedded watermark; Reliable detection
• CPC Class Codes: G06T 2201/0083 whereby only watermarked image required at decoder, e.g. source-based, blind, oblivious

Abstract

An improved method and computer program product for detecting an amount of rotation or magnification in a modified image, of the type including the steps of: a) embedding a marker image having a pair of identical features separated by a distance d and oriented at an angle α in an original image to produce a marked image, the marked image having been rotated and/or magnified to produce the modified image; b) performing an autocorrelation on the modified image to produce a pair of autocorrelation peaks corresponding to the location of the features of the marker image in the modified image; and c) comparing the separation d and orientation α of the autocorrelation peaks with the separation d and orientation α of the features in the marker image to determine the amount of rotation and magnification in the modified image, wherein the improvement comprises the step of: d) preprocessing the modified image to have a constant standard deviation over all local regions in the image.

Claims

1. An improved method for detecting an amount of rotation or magnification in a modified image, of the type including the steps of:

a) embedding a marker image having a pair of identical features separated by a distance d and oriented at an angle α

in an original image to produce a marked image, the marked image having been rotated and/or magnified to produce the modified image;

b) performing an autocorrelation on the modified image to produce a pair of autocorrelation peaks corresponding to the location of the features of the marker image in the modified image; and

c) comparing the separation d and orientation α

of the autocorrelation peaks with the separation d and orientation α

of the features in the marker image to determine the amount of rotation and magnification in the modified image, wherein the improvement comprises the step of;

d) preprocessing the modified image to have a constant standard deviation over all local regions in the image;

wherein the preprocessing step comprises the steps of;

a) specifying a desired standard deviation;

b) for each pixel within the modified image,

i) calculating a local standard deviation, and if calculated standard deviation is less than the desired standard deviation, replacing it with a random number generated from a process having the same standard deviation as the desired standard deviation,

ii) multiplying the pixel value by the ratio of the desired standard deviation and the local standard deviation to form a modified pixed value; and

c) replacing all of the pixel values in the modified image with modified pixel values.

2. The improved method claimed in claim 1, wherein filtering operations and the autocorrelation are performed in the frequency domain.

3. The improved method claimed in claim 1, further comprises the step of processing the autocorrelation with a Wiener filter to improve the signal to noise ratio of the autocorrelation peaks.

4. An improved computer program product, including a computer readable storage medium having a computer program stored thereon for detecting an amount of rotation or magnification in a modified image, for performing the steps of:

a) embedding a marker image having a pair of identical features separated by a distance d and oriented at an angle α

in an original image to produce a marked image, the marked image having been rotated and/or magnified to produce the modified image;

b) performing an autocorrelation on the modified image to produce a pair of autocorrelation peaks corresponding to the location of the features of the marker image in the modified image; and

c) comparing the separation d′ and

orientation α

′

of the autocorrelation peaks with the separation d and orientation α

of the features in the marker image to determine the amount of rotation and magnification in the modified image, wherein the improvement comprises the step of preprocessing the modified image to have a constant standard deviation over all local regions in the image; and

wherein the preprocessing step comprises the steps of;

a) specifying a desired standard deviation;

b) for each pixel within the modified image,

i) calculating a local standard deviation, and if calculated standard deviation is less than the desired standard deviation replacing it with a random number generated from a process having the same standard deviation as the desired standard deviation,

ii) multiplying the pixel value by the ratio of the desired standard deviation and the local standard deviation to form a modified pixel value; and

c) replacing all of the pixel values in the modified image with modified pixel values.

5. The improved computer program product claimed in claim 4, wherein filtering operations and the autocorrelation are performed in the frequency domain.

6. The computer program product method claimed in claim 5, further comprising the step of processing the autocorrelation with a Wiener filter to improve the signal to noise ratio of the autocorrelation peaks.