Image Sign Classifier

• US 20140294291A1
• Filed: 03/26/2013
• Published: 10/02/2014
• Est. Priority Date: 03/26/2013
• Status: Active Grant

• Current Assignee: Micro Focus LLC (Open Text Corporation)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Zhang, Tong
• Granted Patent: US 9,092,696 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: G06F 18/40 Software arrangements specially adapted for pattern recognition, e.g. user interfaces or toolboxes therefor
• CPC Class Codes: G06V 10/774 Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
• CPC Class Codes: G06V 10/945 User interactive design; Environments; Toolboxes
• CPC Class Codes: G06V 20/582 of traffic signs

Abstract

Examples disclosed herein relate to an image sign classifier. In one implementation, a processor causes a user interface to be displayed to receive information related to a target sign type in an image. The processor may train an image sign classifier based on the information to recognize the target sign type and output information related to the trained classifier.

Claims

1. An apparatus, comprising:

a processor to;

cause a user interface to be displayed to receive color and shape information related to a target sign type in an image and to receive a training data set including images of the target sign type;

extract feature information from the training data set;

train an image sign classifier based on the color, shape, and feature information to recognize the target sign type; and

output information related to the trained classifier.

2. The apparatus of claim 1, wherein the user interface is further to receive information about preprocessing an image and wherein training the image classifier comprises preprocessing the images according to the preprocess information.

3. The apparatus of claim 1, wherein the user interface receives at least one of:

information about a method to classify image color information, information about a method to classify shape information, and information about a method to classify feature information.

4. The apparatus of claim 1, wherein the processor is further to use the extracted feature information to distinguish between classes of signs of the particular sign type.

5. A method, comprising:

cause a user interface to be displayed to receive color and shape information related to a particular sign type and a training data set of images including the particular sign type;

extract feature information related to the particular sign type from the training data set;

train an image classifier to detect the particular sign type based on the color, shape, and feature information; and

output information related to the trained classifier.

6. The method of claim 5, further comprising applying the trained classifier to a new image to determine if the image includes the particular type of sign.

7. The method of claim 5, wherein the user interface further receives information about preprocessing images to be classified.

8. The method of claim 5, wherein receiving color information comprises causing the user interface to allow a user to select a portion of a color on at least one of the received images to indicate a color of the particular type of sign.

9. The method of claim 5, wherein receiving shape information comprises causing the user interface to allow a user to select a shape from a shape library to indicate a shape of the particular type of sign.

10. The method of claim 5, comprising training the classifier to determine different classes of the particular sign type based on the features extracted from the received images.

11. The method of claim 5, wherein the user interface further receives information related to a method to performing at least one of:

color detection, shape detection, and feature detection.

12. A machine-readable non-transitory storage medium comprising instructions executable by a processor to:

create a target sign type image detection classifier based on user input received from a user interface related to the target sign color and shape information;

determine feature information related to different classes of the target sign type based on a training data set including images of the target sign type;

train an image classifier to detect the target sign type and distinguish between classes of the target sign type based on the image feature; and

output information related to the image classifier.

13. The machine-readable non-transitory storage medium of claim 12, wherein instructions to train the image classifier comprise instructions to train the image classifier based on user input related to a classifier training method.

14. The machine-readable non-transitory storage medium of claim 12, further comprising instructions to:

receive an image;

apply the image classifier to determine if the received image includes the target sign type; and

determine a class of the target sign type in the image if the target sign type is included.

15. The machine-readable non-transitory storage medium of claim 12, further comprising instructions to create the target sign type image detection classifier based on user input related to preprocessing image information.