METHOD AND APPARATUS FOR GENERATING TEXT LINE CLASSIFIER

• US 20160283814A1
• Filed: 03/24/2016
• Published: 09/29/2016
• Est. Priority Date: 03/25/2015
• Status: Active Grant

• Current Assignee: Alibaba Group Holding Ltd.
• Original Assignee: Alibaba Group Holding Ltd.
• Inventors: JIN, Xuan, WANG, Tianzhou, XUE, Qin
• Granted Patent: US 10,146,994 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 18/00 Pattern recognition
• CPC Class Codes: G06F 18/28 Determining representative reference patterns, e.g. by averaging or distorting; Generating dictionaries
• CPC Class Codes: G06V 30/1914 Determining representative reference patterns, e.g. averaging or distorting patterns; Generating dictionaries, e.g. user dictionaries
• CPC Class Codes: G06V 30/2268 using stroke segmentation
• CPC Class Codes: G06V 30/2445 Alphabet recognition, e.g. Latin, Kanji or Katakana
• CPC Class Codes: G06V 30/293 of characters other than Kanji, Hiragana or Katakana
• CPC Class Codes: G06V 30/413 Classification of content, e.g. text, photographs or tables

Abstract

A method of generating a text line classifier including generating text line samples by use of a present terminal system font reservoir. The method also includes extracting features from the text line samples and pre-stored marked-up samples. The method further includes training models by use of the extracted features to generate a text line classifier for recognizing text regions. With the system font reservoir being utilized for generating text line samples, the generated text line classifiers can target different scenes or different requirements for text region recognition with a high degree of applicability and wide application in addition to ease of implementation. Together with the combinational use of the marked up samples for extracting features from the text line samples, the generated text line classifiers provide for enhanced classification efficiency and accuracy.

Claims

1. A method of generating a text line classifier, the method comprising:

generating text line samples using a present terminal system font reservoir;

extracting features from the text line samples and pre-stored marked-up samples; and

training models using the extracted features to generate a text line classifier for recognizing text regions.

2. The method of claim 1, further comprising:

detecting an image to be recognized to obtain a detection result;

generating a score using the generated text line classifier for the detection result;

determining that the image is a text region if the score is greater than a pre-determined threshold; and

determining that the image is a non-text region if the score is less than the pre-determined threshold.

3. The method of claim 1, wherein the generating text line samples comprises:

generating character samples using the present terminal system font reservoir; and

processing the character samples to generate text line samples of different types, wherein a text line sample comprises character samples of;

a same size;

a same rotation angle; and

a same font, and wherein a text line sample has a percentage of commonly used characters greater than a pre-determined threshold.

4. The method of claim 1, wherein the extracting features from the text line samples and pre-stored marked-up samples comprises:

extracting, from images corresponding to the text line samples, one or more of a gradient orientation histogram feature, a gradient magnitude histogram feature, a pixel histogram feature, and a pixel histogram change feature;

obtaining continuous regions of the text line samples and the marked-up samples; and

extracting features of the continuous regions.

5. The method of claim 1, wherein the training models comprises:

generating models corresponding to types of the text line samples based on the extracted features; and

assigning weights to the models based on the marked up samples to generate a text line classifier.

6. The method of claim 4, wherein the obtaining continuous regions of the marked-up samples comprises:

obtaining continuous regions using a first estimation algorithm, the first estimation algorithm comprising a maximal stable extremal region (MSER) algorithm and an algorithm based upon the MSER algorithm.

7. The method of claim 4, wherein the extracting features of the continuous regions further comprises:

utilizing a second estimation algorithm to obtain stroke width features for the continuous region, wherein the second estimation algorithm comprises a stroke width transform (SWT) algorithm and a stroke feature transform (SFT) algorithm.

8. An apparatus for generating a text line classifier, the apparatus comprising:

a processor; and

a non-transitory computer-readable medium coupled to the processor, the non-transitory computer-readable medium having computer-readable instructions stored thereon to be executed by the processor, the instructions comprising;

a generating module configured to generate text line samples using a present terminal system font reservoir;

an extracting module configured to extract features from the text line samples and pre-stored marked-up samples; and

a training module configured to train models using the extracted features to generate a text line classifier for recognizing text regions.

9. The apparatus of claim 8, further comprising:

a detecting module configured to detect an image to be recognized to obtain a detection result; and

a recognizing module configured to generate a score using the generated text line classifier for the detection result, wherein the recognition module determines that the image is a text region if the score is greater than a pre-determined threshold; and

wherein the recognition module determines that the image is a non-text region if the score is less than the pre-determined threshold.

10. The apparatus of claim 8, wherein the generating module is configured to generate character samples using the present terminal system font reservoir and configured to process the character samples to generate text line samples of different types, and wherein a text line sample comprises character samples of:

a same size;

a same rotation angle; and

a same font, and wherein a text line sample has a percentage of commonly used characters greater than a pre-determined threshold.

11. The apparatus of claim 8, wherein the extracting module comprises:

a first extracting module configured to extract, from images corresponding to the text line samples, one or more of a gradient orientation histogram feature, a gradient magnitude histogram feature, a pixel histogram feature, and a pixel histogram change feature; and

a second extracting module configured to obtain continuous regions of the text line samples and the marked-up samples and to extract features of the continuous regions.

12. The apparatus of claim 8, wherein the training module is configured to generate models corresponding to types of the text line samples based on the extracted features and to assign weights to the models based on the marked up samples to generate a text line classifier.

13. A non-transitory computer readable storage medium having embedded therein program instructions, when executed by one or more processors of a device, causes the device to execute a process for generating a text line classifier, the process comprising:

generating text line samples using a present terminal system font reservoir;

extracting features from the text line samples and pre-stored marked-up samples; and

training models using the extracted features to generate a text line classifier for recognizing text regions.

14. The non-transitory computer readable storage medium of claim 13, wherein the process further comprises:

detecting an image to be recognized to obtain a detection result;

generating a score using the generated text line classifier for the detection result;

determining that the image is a text region if the score is greater than a pre-determined threshold; and

determining that the image is a non-text region if the score is less than the pre-determined threshold.

15. The non-transitory computer readable storage medium of claim 13, wherein the generating text line samples comprises:

generating character samples using the present terminal system font reservoir; and

processing the character samples to generate text line samples of different types, wherein a text line sample comprises character samples of;

a same size;

a same rotation angle; and

a same font, and wherein a text line sample has a percentage of commonly used characters greater than a pre-determined threshold.

16. The non-transitory computer readable storage medium of claim 13, wherein the extracting features from the text line samples and pre-stored marked-up samples comprises:

extracting, from images corresponding to the text line samples, one or more of a gradient orientation histogram feature, a gradient magnitude histogram feature, a pixel histogram feature, and a pixel histogram change feature;

obtaining continuous regions of the text line samples and the marked-up samples; and

extracting features of the continuous regions.

17. The non-transitory computer readable storage medium of claim 13, wherein the training models comprises:

generating models corresponding to types of the text line samples based on the extracted features; and

assigning weights to the models based on the marked up samples to generate a text line classifier.

18. The non-transitory computer readable storage medium of claim 13, wherein the obtaining continuous regions of the marked-up samples comprises:

obtaining continuous regions by a first estimation algorithm, the first estimation algorithm comprising a maximal stable extremal region (MSER) algorithm and an algorithm based upon the MSER algorithm.

19. The non-transitory computer readable storage medium of claim 13, wherein the extracting features of the continuous regions comprises:

utilizing a second estimation algorithm to obtain stroke width features for the continuous region, wherein the second estimation algorithm comprises a stroke width transform (SWT) algorithm and a stroke feature transform (SFT) algorithm.