Method of directed pattern enhancement for flexible recognition

• US 20070127834A1
• Filed: 12/07/2005
• Published: 06/07/2007
• Est. Priority Date: 12/07/2005
• Status: Active Grant

• Current Assignee: Leica Microsystems CMS GmbH (Danaher Corporation)
• Original Assignee: Shih-Jong Lee
• Inventors: Oh, Seho, Lee, Shih-Jong
• Granted Patent: US 8,014,590 B2
• Time in Patent Office: Days
• US Class Current: 382/254
• CPC Class Codes: G06F 18/231 Hierarchical techniques, i.e. dividing or merging pattern sets so as to obtain a dendrogram
• CPC Class Codes: G06F 18/2415 based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
• CPC Class Codes: G06V 10/28 Quantising the image, e.g. histogram thresholding for discrimination between background and foreground patterns
• CPC Class Codes: G06V 10/7625 Hierarchical techniques, i.e. dividing or merging patterns to obtain a tree-like representation; Dendograms
• CPC Class Codes: G06V 10/764 using classification, e.g. of video objects

Abstract

A directed pattern enhancement method receives a learning image and pattern enhancement directive. Pattern enhancement learning is performed using the learning image and the pattern enhancement directive to generate pattern enhancement recipe. An application image is received and a pattern enhancement application is performed using the application image and the pattern enhancement recipe to generate pattern enhanced image. A recognition thresholding is performed using the pattern enhanced image to generate recognition result. The pattern enhancement directive consists of background directive, patterns to enhance directive, and patterns to suppress directive. A partitioned modeling method receives an image region and performs feature extraction on the image region to generate characterization feature. A hierarchical partitioning is performed using the characterization feature to generate hierarchical partitions. A model generation is performed using the hierarchical partitions to generate partition model. The partitioned modeling further performs a partitioned matching step that matches an input point to the partition model to generate a matching score output. A partition model update method receives a partition model and input data for model update. A partition model update is performed using the partition model and the data to generate an updated partition model.

Claims

1. A partitioned modeling method comprising the steps of:

a) Input image region;

b) Perform feature extraction on the image region having characterization feature output;

c) Perform hierarchical partitioning using the characterization feature having hierarchical partitions output;

d) Perform model generation using the hierarchical partitions having partition model output.

2. The partitioned modeling method of claim 1 wherein the hierarchical partitioning method generates multiple points pre-clusters at the bottom of the hierarchy.

3. The partitioned modeling method of claim 1 wherein the model generation method estimates the model parameters for the partitions from the lowest level to the highest level using incremental update rule.

4. The partitioned modeling method of claim 1 further comprises a partitioned matching step that matches an input point to the partition model having a matching score output.

5. The partitioned matching step of claim 4 further comprises a filtering step to remove noise on matching score.

8. The partition model update method of claim 5 wherein the partition model update performs regressive update comprising the steps of:

a) Perform regressive parameter update using partition model and the data having regressive updated parameters output;

b) Perform regressive model update using the regressive updated parameters having regressive updated partition model output.

6. A partition model update method comprising the steps of:

a) Input a partition model;

b) Input data for model update;

c) Perform partition model update using the partition model and the data having a updated partition model output.

7. The partition model update method of claim 6 wherein the partition model update performs progressive update comprising the steps of:

a) Perform uniqueness detection using partition model and the data having a unique set and a similar set output;

b) Perform supplementary model generation using the unique set having supplementary model output;

c) Perform combined model generation using the supplementary model and the partition model having combined model output;

d) Perform progressive parameter update using the combined model and the a similar set having progressive updated partition model output.

9. A directed pattern enhancement method comprising the steps of:

a) Input learning image and pattern enhancement directive;

b) Perform pattern enhancement learning using the learning image and the pattern enhancement directive having pattern enhancement recipe output;

c) Input application image;

d) Perform pattern enhancement application using the application image and the pattern enhancement recipe having pattern enhanced image output.

10. The directed pattern enhancement method of claim 9 further performs recognition thresholding using the having pattern enhanced image having recognition result output.

11. The directed pattern enhancement method of claim 9 wherein the pattern enhancement directive specifies from a set consisting of

a) Background directive,

b) Patterns to enhance directive,

c) Patterns to suppress directive.

12. The directed pattern enhancement method of claim 11 wherein the pattern enhancement learning comprising the steps of:

a) Perform background learning using the background directive and the learning image having background recipe output;

b) Perform foreground pattern extraction using pattern enhancement directive the learning image and the background recipe having foreground pattern output;

c) Perform foreground pattern learning using the pattern enhancement directive the learning image and the foreground pattern having foreground pattern recipe output.

13. The directed pattern enhancement method of claim 9 wherein the pattern enhancement application comprising the steps of:

a) Perform background matching using the background recipe and the application image having non-background or background score output;

b) Perform foreground enhancement pattern matching using foreground enhancement pattern recipe and the application image having enhancement score output;

c) Perform integration using the enhancement score and the background score having pattern enhanced image output.

14. The pattern enhancement application of claim 13 further comprising the steps of:

a) Perform foreground suppression pattern matching using foreground suppression pattern recipe and the application image having suppression score output;

b) Perform integration using the enhancement score, the suppression score and the background score having pattern enhanced image output.

15. The integration step of claim 14 further comprising a special integration step when both the enhancement score and suppression score have high values.

16. The special integration step of claim 15 uses the average enhancement score and average suppression score of the neighbors of the points.

17. The special integration step of claim 15 uses whole neighbor as template to create a new enhancement score and a new suppression score.

18. An update learning method comprising the steps of:

a) Input image and pattern enhancement recipe;

b) Perform pattern enhancement application using the pattern enhancement recipe and the image having pattern enhanced image output;

c) Perform performance review and directive generation using the pattern enhanced image having progressive pattern enhancement directive output;

d) Perform pattern enhancement progressive update learning using the image and pattern enhancement recipe having progressive updated pattern enhancement recipe.

19. The update learning method of claim 18 wherein the progressive update learning updates background recipe foreground enhancement pattern recipe and foreground suppression pattern recipe.

20. The update learning method of claim 18 further comprises a regressive update learning step using pattern enhancement recipe, learned image and regressive recognition enhancement directive having regressive updated pattern enhancement recipe output.