Method for measuring and analyzing digital video quality

• US 20030031368A1
• Filed: 08/01/2001
• Published: 02/13/2003
• Est. Priority Date: 08/01/2001
• Status: Active Grant

• Current Assignee: Geo Semiconductor Incorporated, University of Central Florida (State University System of Florida)
• Original Assignee: Teranex, Inc. (Blackmagic Design Pty Limited), University of Central Florida (State University System of Florida)
• Inventors: Myler, Harley R., Dyke-Lewis, Michele Van
• Granted Patent: US 6,577,764 B2
• Time in Patent Office: Days
• US Class Current: 382/228
• CPC Class Codes: G06T 7/0004 Industrial image inspection
• CPC Class Codes: H04N 17/004 for digital television systems
• CPC Class Codes: H04N 17/02 for colour television signals

Abstract

A method and system for analyzing and measuring image quality between two images. A series of conversions and transformations of image information are preformed to produce a single measure of quality. A YCrCb frame sequence is first converted using RGB conversion to an RGB frame sequence. The resulting RGB frame sequence is converted using spherical coordinate transform (SCT) conversion to SCT images. A Gabor filter is applied to the SCT images to produce a Gabor Feature Set, and a statistics calculation is applied to the Gabor Feature Set. The resulting Gabor Feature Set statistics are produced for both the reference frame and the frame to be compared. Quality is computed for these Gabor Feature Set statistics to produce a video quality measure. Spectral decomposition of the frames is also performable for the Gabor Feature Set, rather than the statistics calculation, allowing graphical comparison of results for the compared frames.

Claims

1. A method for comparing image quality for a first image and a second image, the method comprising:

identifying a first set of Red-Green-Blue (RGB) image data for the first image and a second set of RGB image data for the second image;

transforming the first RGB image data to spherical coordinates, wherein the spherical coordinates for the first RGB image data comprise a first set of Spherical Coordinate Transform (SCT) image data;

transforming the second RGB image data to spherical coordinates, wherein the spherical coordinates for the first RGB image data comprises a second set of SCT image data;

applying a Gabor filter to the first set of SCT image data to produce a first Gabor feature set;

applying the Gabor filter to the second set of SCT image data to produce a second Gabor feature set;

applying a statistics calculation to the first Gabor feature set to produce a first set of Gabor feature statistics;

applying the statistics calculation to the second Gabor feature set to produce a second set of Gabor feature statistics; and

comparing the first set of Gabor feature statistics to the second set of Gabor feature statistics.

2. The method of claim 1, wherein identifying a first RGB image for the first image and a second RGB image for the second image further comprises:

converting a first luma-red chroma-blue chroma (YCrCb) image for the first image to the first RGB image; and

converting a second YCrCb image for the second image to the second RGB image.

3. The method of claim 1, wherein comparing the first set of Gabor feature statistics to the second set of Gabor feature statistics comprises:

computing a quality measure for the first set of Gabor feature statistics compared to the second set of Gabor feature statistics.

4. The method of claim 3, wherein the first set of Gabor feature statistics includes a first plurality of Gabor feature numbers and a first plurality of values, each of the first plurality of values corresponding to one from the first plurality of Gabor feature numbers;

wherein the second set of Gabor feature statistics includes a second plurality of Gabor feature numbers and a second plurality of values, each of the second plurality of values corresponding to one from the second plurality of Gabor feature numbers;

wherein each of the first plurality of Gabor feature numbers has a corresponding one of the second plurality of Gabor feature numbers; and

wherein computing a quality measure for the first set of Gabor feature statistics compared to the second set of Gabor feature statistics includes;

for each of the values for each of the first plurality of Gabor feature numbers determining a difference with the corresponding value for each of the second plurality of Gabor feature numbers, such that a plurality of differences are produced; and

summing the plurality of produced differences.

5. The method of claim 3, wherein the first set of Gabor feature statistics includes a first plurality of Gabor feature numbers and a corresponding first plurality of values for the first plurality of Gabor feature numbers;

wherein the second set of Gabor feature statistics includes a second plurality of Gabor feature numbers and a corresponding second plurality of values for the second plurality of Gabor feature numbers;

wherein each of the first plurality of Gabor feature numbers has a corresponding one of the second plurality of Gabor feature numbers; and

wherein computing a quality measure for the first set of Gabor feature statistics compared to the second set of Gabor feature statistics includes;

summing the corresponding values for each of the first plurality of Gabor feature numbers to produce a first sum;

summing the corresponding values for each of the second plurality of Gabor feature numbers to produce a second sum; and

determining the difference between the first sum and the second sum.

6. The method of claim 4, further comprising:

normalizing the summed plurality of produced differences.

7. The method of claim 6, wherein the first plurality of Gabor feature numbers comprise a first discrete number of first Gabor feature numbers;

and

wherein normalizing the summed plurality of produced differences comprises;

dividing the summed plurality of produced differences by the first discrete number of first Gabor feature numbers.

8. The method of claim 7, wherein the second plurality of Gabor feature numbers comprise a second discrete number of second Gabor feature numbers;

and

wherein the first discrete number of first Gabor feature numbers is equal to the second discrete number of second Gabor feature numbers.

9. The method of claim 1, wherein identifying a first RGB image for the first image comprises:

determining if the first image comprises an RGB image.

10. The method of claim 9, wherein identifying a first set of RGB image data for the first image and a second set of RGB image data for the second image comprises:

determining if the first image comprises RGB image data; and

determining if the second image comprises RGB image data.

11. The method of claim 1, wherein comparing the first set of Gabor feature statistics to the second set of Gabor feature statistics comprises:

producing a spectral decomposition for each of the first Gabor feature set and the second Gabor feature set.

12. The method of claim 4, further comprising:

producing a spectral decomposition for each of the first Gabor feature set and the second Gabor feature set.

13. The method of claim 12, wherein producing a spectral decomposition for each of the first Gabor feature set and the second Gabor feature set further comprises:

for each of the first plurality of Gabor feature numbers, graphically depicting the corresponding one of the first plurality of values; and

for each of the second plurality of Gabor feature numbers, graphically depicting the corresponding one of the second plurality of values.

14. A method for analyzing quality of an image, comprising:

identifying Red-Green-Blue (RGB) image data for the image;

transforming the RGB image data to spherical coordinates, such that the image includes Spherical Coordinate Transform (SCT) image data;

applying a Gabor filter to the SCT image data to produce a Gabor feature set for the image; and

applying a statistics calculation to the Gabor feature set for the image to produce Gabor feature statistics for the image.

15. A method for comparing image quality for a first image and a second image, the method comprising:

identifying a first set of Red-Green-Blue (RGB) image data for the first image and a second set of RGB image data for the second image;

transforming the first RGB image data to spherical coordinates, wherein the spherical coordinates for the first RGB image data comprise a first set of Spherical Coordinate Transform (SCT) image data;

transforming the second RGB image data to spherical coordinates, wherein the spherical coordinates for the first RGB image data comprises a second set of SCT image data;

applying a Gabor filter to the first set of SCT image data to produce a first Gabor feature set;

applying the Gabor filter to the second set of SCT image data to produce a second Gabor feature set; and

producing a spectral decomposition for each of the first Gabor feature set and the second Gabor feature set.

16. A system for comparing image quality for a first image and a second image, the system comprising:

a processor for receiving and processing the first image and the second image; and

a repository for storing information from the received first image and the received second image;

wherein the processor identifies a first set of Red-Green-Blue (RGB) image data for the first image and a second set of RGB image data for the second image;

wherein the processor transforms the first RGB image data to spherical coordinates, wherein the spherical coordinates for the first RGB image data comprise a first set of Spherical Coordinate Transform (SCT) image data;

wherein the processor transforms the second RGB image data to spherical coordinates, wherein the spherical coordinates for the second RGB image data comprise a second set of SCT image data;

wherein the processor applies a Gabor filter to the first set of SCT image data to produce a first Gabor feature set;

wherein the processor applies a Gabor filter to the second set of SCT image data to produce a second Gabor feature set;

wherein the processor applies a statistics calculation to the first Gabor feature set to produce a first set of Gabor feature statistics;

wherein the processor applies the statistics calculation to the second Gabor feature set to produce a second set of Gabor feature statistics; and

wherein the processor compares the first set of Gabor feature statistics to the second set of Gabor feature statistics.

17. The system of claim 16, wherein the repository comprises a database.

18. The system of claim 16, wherein the processor comprises a terminal.

19. The system of claim 18, wherein the terminal comprises one selected from a group consisting of a personal computer, a minicomputer, a microcomputer, and a main frame computer.

20. A system for comparing image quality for a first image and a second image, the system comprising:

means for identifying a first set of Red-Green-Blue (RGB) image data for the first image and a second set of RGB image data for the second image;

means for transforming the first RGB image data to spherical coordinates, wherein the spherical coordinates for the first RGB image data comprise a first set of Spherical Coordinate Transform (SCT) image data;

means for transforming the second RGB image data to spherical coordinates, wherein the spherical coordinates for the first RGB image data comprises a second set of SCT image data;

means for applying a Gabor filter to the first set of SCT image data to produce a first Gabor feature set;

means for applying the Gabor filter to the second set of SCT image data to produce a second Gabor feature set;

means for applying a statistics calculation to the first Gabor feature set to produce a first set of Gabor feature statistics;

means for applying the statistics calculation to the second Gabor feature set to produce a second set of Gabor feature statistics; and

means for comparing the first set of Gabor feature statistics to the second set of Gabor feature statistics.