Determining regions of interest in synthetic images

• US 7,848,567 B2
• Filed: 09/23/2004
• Issued: 12/07/2010
• Est. Priority Date: 09/23/2004
• Status: Expired due to Fees

• Current Assignee: Fujifilm Business Innovation Corp. (Fujifilm Holdings Corporation)
• Original Assignee: Fuji Xerox Company Limited (Xerox Holdings Corp.)
• Inventors: Girgensohn, Andreas, Chiu, Patrick, Liu, Qiong
• Primary Examiner(s): Chen; Wenpeng
• Assistant Examiner(s): Thirugnanam; Gandhi
• Application Number:
  US10/948,823
• Publication Number:
  US 20060062456A1
• Time in Patent Office: 2,266 Days
• Field of Search: 382/162, 382/164, 382/165, 382/173, 382/175, 382/180, 382190-191, 358/2.1, 358/453, 358/538
• US Class Current: 382/164
• CPC Class Codes: G06V 10/25 Determination of region of interest [ROI] or a volume of interest [VOI]

Abstract

An algorithm for finding regions of interest (ROI) in synthetic images based on an information driven approach in which sub-blocks of a set of synthetic image are analyzed for information content or compressibility based on textural and color features. A DCT may be used to analyze the textural features of a set of images and a color histogram may be used to analyze the color features of the set of images. Sub-blocks of low compressibility are grouped into ROIs using a type of morphological technique. Unlike other algorithms that are geared for highly specific types of ROI (e.g. OCR text detection), the method of the present invention is generally applicable to arbitrary synthetic images. The present invention can be used with several other image applications, including Stained-Glass collages presentations.

Claims

1. A method for finding regions of interest in a synthetic image comprising:

generating a plurality of sub-blocks for a plurality of synthetic images;

computing a texture value derived from one or more color features associated with the plurality of sub-blocks, the texture value computed on at least one of one or more color bands for each of the sub-blocks, the texture value being computed using discrete cosine transform;

computing a color value associated with the color value of each of the plurality of sub-blocks for one or more of the color bands, wherein the color value includes computing a color correlogram on one of one or more of the color bands for each of the sub-blocks in each of the synthetic images;

determining an information value based on a weighted average of the texture value and the color value of each of the plurality of sub-blocks; and

automatically grouping the sub-blocks based on the informational value to form a region of interest, wherein grouping includes grouping neighboring sub-blocks together that have an information value that resides in a predetermined range, wherein the grouped neighboring sub-blocks have a density that does not exceed an average density of the plurality of sub-blocks.

2. The method of claim 1 wherein grouping the sub-blocks includes:

binarizing the information value for each sub-block of each synthetic image.

3. The method of claim 2 wherein binarizing includes assigning a value of one to a sub-block with an information value above a standard deviation.

4. The method of claim 1 wherein the information value is a binarized value, and the predetermined range includes sub-blocks having an information value of one.

5. The method of claim 1 wherein standard morphology operations of dilation and erosion can be used in conjunction with the grouping.

6. The method of claim 1 wherein determining a dominant group to be the largest group of sub-blocks.

7. The method of claim 1 wherein a user may supplement the automatic grouping and select one or more groups of sub-blocks as regions of interest.

8. The method of claim 1 further comprising:

selecting a number of regions of interest from the plurality of synthetic images using uniform sub-sampling of the synthetic images.

9. The method of claim 1 further comprising:

selecting a number of regions of interest from the plurality of synthetic images by using a genetic segmentation algorithm.

10. A non-transistory computer readable storage medium having instructions stored thereon that when processed by one or more processors cause a system to:

generate a plurality of sub-blocks for a plurality of synthetic images, wherein the sub-blocks include two or more pixels;

compute a texture value derived from one or more color features associated with the plurality of sub-blocks, the texture value computed on at least one of one or more color bands for each of the sub-blocks, the texture value being computed using an algorithm selected from the group consisting of a discrete cosine transform, a wavelet and a Haar decomposition;

compute a color value associated with the color value of each of the plurality of sub-blocks for one or more of the color bands;

determine an information value based on a weighted average of the texture value and the color value; and

group the sub-blocks based on the informational value to form a region of interest, wherein the grouping includes grouping neighboring sub-blocks together that have an information value that resides in a predetermined range, wherein the grouped neighboring sub-blocks have a density that does not exceed an average density of the plurality of sub-blocks.

11. The computer readable storage medium of claim 10 wherein grouping the sub-blocks includes:

binarizing the information value for each sub-block of each synthetic image.

12. The computer readable storage medium of claim 11 wherein binarizing includes assigning a value of one to a sub-block with an information value above a standard deviation.

13. The computer readable storage medium of claim 10 wherein the information value is a binarized value, and the predetermined range includes sub-blocks having an information value of one.

14. The computer readable storage medium of claim 10 wherein standard morphology operations of dilation and erosion can be used in conjunction with the grouping the neighboring sub-blocks.

15. The computer readable storage medium of claim 10 wherein the instructions cause a system to determine a dominant group to be the largest group of sub-blocks.

16. The computer readable storage medium of claim 10 wherein the instructions allow a user to select one or more groups of sub-blocks as regions of interest.

17. The computer readable storage medium of claim 10 wherein computing a color value associated with a plurality of sub-blocks includes:

computing a color histogram on three color bands for each of the sub-blocks in each of the synthetic images.

18. The computer readable storage medium of claim 10 wherein computing a color value includes:

computing a color correlogram on three color bands for each of the sub-blocks in each of the synthetic images.

19. The computer readable storage medium of claim 10 wherein the instructions cause the system to:

compute a textural value using spectral decomposition on the plurality of sub-blocks for the plurality of synthetic images, the spectral decomposition being computed on three color bands for each of the plurality of sub-blocks;

compute a weighted-sum value by combining the color value with the textural value; and

determine an information value associated with each of the plurality of sub-blocks based on the weighted-sum value.

20. The computer readable storage medium of claim 10 wherein the instructions further cause the system to select a number of regions of interest from the plurality of synthetic images using uniform sub-sampling of the synthetic images.

21. The computer readable storage medium of claim 10 wherein the instructions further cause the system to select a number of regions of interest from the plurality of synthetic images by determining the distance function between the color histogram for each of the regions of interest and the average color profile for the plurality of synthetic images.

22. The computer readable storage medium of claim 10 wherein the instructions further cause the system to select a number of regions of interest from the plurality of synthetic images by using a genetic segmentation algorithm.