Image subregion querying using color correlograms

• US 6,430,312 B1
• Filed: 12/28/1998
• Issued: 08/06/2002
• Est. Priority Date: 12/29/1997
• Status: Expired due to Fees

• Current Assignee: Cornell Research Foundation Incorporated (Cornell University)
• Original Assignee: Cornell Research Foundation Incorporated (Cornell University)
• Inventors: Zhu, Wei-Jing, Mitra, Mandar, Kumar, Shanmugasundaram Ravi, Huang, Jing
• Primary Examiner(s): Tran, Phuoc
• Assistant Examiner(s): Alavi, Amir
• Application Number:
  US09/221,473
• Time in Patent Office: 1,317 Days
• Field of Search: 382/162-172, 382/173, 382/191-198, 382/201, 382/205, 382/209, 382/278, 382/285, 382/293, 382/295, 382/305, 382/309
• US Class Current: 382/165
• CPC Class Codes: G06F 16/58 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: G06F 16/5838 using colour
• CPC Class Codes: G06T 2207/10024 Color image
• CPC Class Codes: G06T 7/90 Determination of colour characteristics
• CPC Class Codes: G06V 10/56 relating to colour

Abstract

A color correlogram (10) is a representation expressing the spatial correlation of color and distance between pixels in a stored image. The color correlogram (10) may be used to distinguish objects in an image as well as between images in a plurality of images. By intersecting a color correlogram of an image object with correlograms of images to be searched, those images which contain the objects are identified by the intersection correlogram.

Claims

1. A computer-implemented method for retrieving an image object from a plurality of images, comprising the steps of:

providing an image object color correlogram;

providing a plurality of color values;

selecting a distance value to be used as the distance between pixels, in the image object and in the plurality of images, to be evaluated for color value;

analyzing said image object according to said color values and said selected distance value;

determining in response to the analyzing step a probability of finding a pixel of a particular color value at said distance value from a pre-selected pixel of a pre-selected color value;

entering said probability into the image object color correlogram;

providing color correlograms for each of said plurality of images; and

intersecting the image object color correlogram with correlograms of the plurality of images to produce an intersection result,

wherein the image object is distinguished by the intersection result from the images which do not contain the image object.

2. The method of claim 1 wherein the intersecting step further comprises comparing a count of a first plurality of color pairs in the image object with a count of a second plurality of color pairs in the image, the first plurality of color pairs correlating to the second plurality of color pairs.

3. The method of claim 1 wherein the step of providing color correlograms for each of said plurality of images further comprises storing the provided correlograms in a database.

4. The method of claim 1 further comprising the steps of:

selecting a plurality of distance values; and

performing said analyzing step, said determining step and said entering step using said plurality of distance values.

5. A system for retrieving an image object from a plurality of images, comprising:

means for providing an image object color correlogram;

means for providing a plurality of color values;

means for selecting a distance value to be used as the distance between pixels to be evaluated for color value in the image object and in the plurality of images;

means for analyzing said image object according to said color values and said selected distance value;

means for determining in response to said means for analyzing, a probability of finding a pixel of a particular color value at said distance value from a pre-selected pixel of a pre-selected color value;

means for entering said probability into said image object color correlogram;

means for providing color correlograms for each of said plurality of images; and

means for intersecting said image object color correlogram with correlograms of the plurality of images to produce an intersection result,

wherein the image object is distinguished from the images which do not contain the image object by the intersection result.

6. The system of claim 5 wherein said means for intersecting further comprises a means for comparing a count of a first plurality of color pairs in the image object with a count of a second plurality of color pairs in the image, the first plurality of color pairs correlating to the second plurality of color pairs.

7. The system of claim 5 further comprising a database for storing said provided correlograms of said plurality of images.

8. The system of claim 5 further comprising:

means for selecting a plurality of distance values;

means for analyzing said image object and said plurality of images according to said color values and said plurality of distance values; and

means for determining, in response to said analyzing means, a probability of finding a pixel of a particular color value for each of said plurality of distance values from a selected pixel of a selected color value.

9. A computer-implemented method of locating an image object in an image comprising the steps of:

providing a plurality of color values and at least one distance value;

computing a color correlogram for the image object using the plurality of color values and the at least one distance value;

computing a color correlogram for the image using the plurality of color values and the at least one distance value;

analyzing the image object and the image to determine a color frequency value for each color value;

assigning the color-frequency value to each pixel in the image object to make a back-projection image object correlogram; and

combining the back-projection image object correlogram with the image correlogram to create a correlogram backprojection indicating the location of the image object in the image.

10. The method of claim 9 wherein

said step of computing the image object color correlogram further comprises computing an autocorrelogram for the image object; and

said step of computing the image color correlogram further comprises computing an autocorrelogram for the image.

11. The method of claim 9 further comprising the step of:

locating the image object by the mathematical center of the image object.

12. The method of claim 9 further comprising the step of:

combining the back-projection image object correlogram with a color histogram of the image object to obtain correction values; and

combining the correction values with the color correlogram of the image to accurately locate the image object.

13. The method of claim 12 further comprising the step of weighting the values of the back-projection image object correlogram and weighting the values of the color histogram to produce weighted correction values to be combined with the color correlogram of the image.

14. A system for locating an image object in an image, comprising:

means for providing a plurality of color values and at least one distance value;

means for computing a color correlogram for the image object using the plurality of color values and the at least one distance value;

means for computing a color correlogram for the image using the plurality of color values and the at least one distance value;

means for analyzing the image object and the image to determine a color frequency value for each color value;

means for assigning the color-frequency value to each pixel in the image object to make a back-projection image object correlogram; and

means for combining the back-projection image object correlogram with the image correlogram to create a correlogram backprojection indicating the location of the image object in the image.

15. The system of claim 14 wherein

said means for computing the image object color correlogram further comprises means for computing an autocorrelogram for the image object; and

said means for computing the image color correlogram further comprises means for computing an autocorrelogram for the image.

16. The system of claim 14 further comprising:

means for combining the back-projection image object correlogram with a color histogram of the image object to obtain correction values; and

means for combining the correction values with the color correlogram of the image to accurately locate the image object.

17. The system of claim 16 further comprising:

means for weighting the values of the back-projection image object correlogram;

means for weighting the values of the color histogram to produce weighted correction values to be combined with the color correlogram of the image.

18. A method for detecting cuts in a sequence of video frames comprising the steps of:

providing an image object;

computing a color correlogram of the image object;

computing a color correlogram of a first video frame;

intersecting the image object color correlogram with the first video frame color correlogram to obtain a first result determining the presence or absence of the image object in the first video frame;

computing a color correlogram of a second video frame, the second video frame being adjacent to the first video frame in the sequence of video frames;

intersecting the image object color correlogram with the second video frame color correlogram to obtain a second result determining the presence or absence of the image object in the second video frame; and

,

comparing the first result with the second result in order to determine a cut between the first video frame and the second video frame,

wherein a cut occurs where the image object is present in one of the adjacent video frames and not present in the other adjacent video frame.