Systems and methods for detecting text

• US 20060222239A1
• Filed: 03/31/2005
• Published: 10/05/2006
• Est. Priority Date: 03/31/2005
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Simard, Patrice Y., Viola, Paul A., Bargeron, David M.
• Granted Patent: US 7,570,816 B2
• Time in Patent Office: Days
• US Class Current: 382/159
• CPC Class Codes: G06F 18/214 Generating training patterns; Bootstrap methods, e.g. bagging or boosting
• CPC Class Codes: G06V 30/19147 Obtaining sets of training patterns; Bootstrap methods, e.g. bagging or boosting
• CPC Class Codes: G06V 30/40 Document-oriented image-based pattern recognition

Abstract

The subject invention relates to facilitating text detection. The invention employs a boosted classifier and a transductive classifier to provide accurate and efficient text detection systems and/or methods. The boosted classifier is trained through features generated from a set of training connected components and labels. The boosted classifier utilizes the features to classify the training connected components, wherein inferred labels are conveyed to a transductive classifier, which generates additional properties. The initial set of features and the properties are utilized to train the transductive classifier. Upon training, the system and/or methods can be utilized to detect text in data under text detection, wherein unlabeled data is received, and connected components are extracted therefrom and utilized to generate corresponding feature vectors, which are employed to classify the connected components using the initial boosted classifier. Inferred labels are utilized to generate properties, which are utilized along with the initial feature vectors to classify each connected component using the transductive classifier.

Claims

1. A system that facilitates detecting text in data, comprising:

an input component that receives data; and

a classification component that automatically detects text in the data via a trained transductive classifier employed in connection with a trained boosted classifier.

2. The system of claim 1, further comprising a connected components identifier that generates a set of connected components from data, the connected components are utilized to generate a set of features that are employed within the boosted classifier and transductive classifier for training and testing.

3. The system of claim 2, the connected components identifier utilizes a union-find algorithm to identify the training connected components.

4. The system of claim 2, the boosted classifier infers labels for the training connected components, the inferred labels are subjected to a clustering process by which feature sets are expanded to define a set of training properties.

5. The system of claim 4, the classification component is trained to detect text with the set of training features and the set of training properties.

6. The system of claim 1, further comprising a feature generator that generates a set of training features that are employed to train the boosted classifier and the transductive classifier.

7. The system of claim 6, the feature generator utilizes one of a spatial relation technique and an activity density technique to generate the set of features.

8. The system of claim 1, the data is one of a document, a spreadsheet, an image, a page, a web page, a video frame, or a drawing.

9. An optical character recognition (OCR) system that employs the system of claim 1 in connection with extracting text from an image.

10. A computer readable medium having stored thereon the components of the system of claim 1.

11. A method for detecting text, comprising:

identifying one or more connected components associated with the unlabeled data under text detection;

utilizing the connected components to extract a feature vector for each connected component;

classifying each connected component represented by its respective feature vector;

employing inferred labels to bin the connected components across a plurality of bins;

computing properties for each bin; and

utilizing the feature vectors and corresponding properties to classify the connected components.

12. The method of claim 11, further comprising utilizing a boosted classifier to classify the connected components represented by respective feature vectors.

13. The method of claim 12, further comprising training the boosted classifier by:

receiving labeled training data that includes connected components, text labels and non-text labels;

extracting a feature vector for each connected component in the training data; and

training the boosted classifier with the labels and feature vectors.

14. The method of claim 13, further comprising utilizing a transductive classifier to classify the connected components with the feature vectors and corresponding computed bin properties.

15. The method of claim 14, further comprising training the transductive classifier by:

employing the boosted classifier to classify each training connected component, represented by a respective feature vector,

utilizing labels inferred from the training connected components to bin the training connected components across a plurality of bins computing properties for each bin; and

utilizing the labels, feature vectors and properties to train the transductive classifier.

16. The method of claim 15, further comprising utilizing one of a spatial relation technique and an activity density technique to generate the feature vectors for the unlabeled data and the labeled training data.

17. The method of claim 15, the bins utilized to delineate the connected components of the unlabeled data and the training data are associated with one or more histograms that correspond to at least one of the following:

a number of pixels in a connected component;

a height of a connected component'"'"'s bounding box;

a width of a connected component'"'"'s bounding box; and

a number of pixels in a connected component divided by a connected component'"'"'s bounding box area.

18. The method of claim 14, the transductive classifier is a second boosted classifier.

19. The method of claim 14, further comprising employing a decision tree with fixed depth in connection with the boosted classifier and the transductive classifier to allow for complex conditional processing of raw features without over-fitting the data.

20. A system for training a text detector, comprising:

means for training a boosted classifier and a transductive classifier with connected components and corresponding text and non-text labels for text detection; and

means for employing the boosted classifier in connection with the transductive classifier to detect text within unlabeled data.