Image concealing via efficient feature selection

• US 8,774,513 B2
• Filed: 01/09/2012
• Issued: 07/08/2014
• Est. Priority Date: 01/09/2012
• Status: Active Grant

• Current Assignee: General Electric Company
• Original Assignee: General Electric Company
• Inventors: Liu, Xiaoming, Tu, Peter Henry, Xue, Ya
• Primary Examiner(s): Tabatabai, Abolfazl
• Application Number:
  US13/346,479
• Publication Number:
  US 20130177244A1
• Time in Patent Office: 911 Days
• Field of Search: 382/103, 382/164, 382/171, 382/173, 382/179, 382/118, 382/159, 382/197, 348/169, 348/267, 711/206, 711/208, 711/209
• US Class Current: 382/173
• CPC Class Codes: G06F 18/2113 by ranking or filtering the set of features, e.g. using a measure of variance or of feature cross-correlation
• CPC Class Codes: G06F 18/2178 based on feedback of a supervisor
• CPC Class Codes: G06F 18/2323 based on graph theory, e.g. minimum spanning trees [MST] or graph cuts
• CPC Class Codes: G06V 10/754 involving a deformation of the sample pattern or of the reference pattern; Elastic matching
• CPC Class Codes: G06V 10/7635 based on graphs, e.g. graph cuts or spectral clustering
• CPC Class Codes: G06V 10/771 Feature selection, e.g. selecting representative features from a multi-dimensional feature space
• CPC Class Codes: G06V 10/7784 based on feedback from supervisors
• CPC Class Codes: G06V 40/171 Local features and components; Facial parts eye characteristics G06V40/18; Occluding parts, e.g. glasses; Geometrical relationships

Abstract

A novel technique for unsupervised feature selection is disclosed. The disclosed methods include automatically selecting a subset of a feature of an image. Additionally, the selection of the subset of features may be incorporated with a congealing algorithm, such as a least-square-based congealing algorithm. By selecting a subset of the feature representation of an image, redundant and/or irrelevant features may be reduced or removed, and the efficiency and accuracy of least-square-based congealing may be improved.

Claims

1. A method comprising:

automatically dividing an initial feature representation of an image into two or more subsets;

selecting at least one representative feature from each of the two or more subsets;

estimating a warping parameter for the image using the at least one representative feature from each of the two or more subsets.

2. The method of claim 1, wherein estimating a warping parameter for the image comprises estimating an initial warping parameter, computing a warping parameter update, and updating the initial warping parameter to calculate a current warping parameter.

3. The method of claim 1, wherein automatically dividing an initial feature representation of an image into two or more subsets comprises generating a graph with features of the image as vertices.

4. The method of claim 3, comprising applying a power iteration clustering algorithm to the graph.

5. The method of claim 1, wherein selecting at least one representative feature from each of the one or more subsets comprises calculating a maximum information compression index for a first feature of the image and a second feature of the image and applying a heuristic algorithm to the image.

6. The method of claim 1, wherein selecting at least one representative feature from each of the one or more subsets comprises calculating a maximum information compression index for a first feature of the image and a second feature of the image and applying a spectral clustering algorithm to the image.

7. The method of claim 6, wherein the predetermined number of subsets is determined automatically.

8. The method of claim 1, wherein automatically dividing an initial feature representation of an image into two or more subsets comprises dividing the initial feature representation of an image into a predetermined number of subsets.

9. A method, comprising:

automatically identifying a first subset of features of a first image, wherein the first image is selected from a plurality of images;

automatically identifying a second subset of features of a second image, wherein the second image is selected from the plurality of images; and

estimating warping parameters for the first image and the second image using the first subset of features and the second subset of features, wherein estimating warping parameters for the first image and the second image using the first subset of features and the second subset of features comprises executing a least-square-based congealing algorithm using the first subset of features and the second subset of features.

10. The method of claim 9, wherein automatically identifying a first subset of features of a first image comprises dividing the first image into two or more subgroups and selecting at least one representative feature from each subgroup.

11. The method of claim 10, comprising automatically predetermining a number of subgroups.

12. The method of claim 10, wherein automatically identifying a first subset of features of a first image comprises generating a graph with features of the first image as vertices and calculating a connectivity between vertices.

13. The method of claim 12, comprising applying a power iteration clustering algorithm to the graph to identify the two or more subgroups.

14. The method of claim 9, wherein estimating warping parameters for the first image and the second image using the first subset of features and the second subset of features comprises calculating an initial warping parameter, computing a warping parameter update, and updating the initial warping parameter with the warping parameter update to calculate a current warping parameter.

15. A method, comprising:

executing an unsupervised feature selection algorithm, comprising;

constructing a graph having features of an image as vertices;

determining a connectivity between the vertices using a maximum information compression index;

partitioning the graph into two or more subsets of features using a power iteration clustering algorithm; and

selecting representative features from each subset of features; and

executing an image congealing algorithm using the subsets of features.

16. The method of claim 15, wherein the power iteration clustering algorithm includes a perturbation enhancement.

17. The method of claim 15, wherein the image congealing algorithm is a least-square-based image congealing algorithm.

18. The method of claim 15, comprising executing multiple iterations of the image congealing algorithm and executing the unsupervised feature selection algorithm before each iteration of the image congealing algorithm.

19. The method of claim 15, wherein a number of the subsets of features is automatically predetermined.