Method for recognizing target object in image, and apparatus

1. A method for recognizing a target object in an image, comprising:

extracting feature data from the image;

transforming the extracted feature data into a uniform expression;

performing automatic clustering for features in the image according to the feature data in the uniform expression and a historical clustering result;

grouping the target object of a known class comprised in an automatic clustering result into the corresponding known class, in order to recognize the target object of the known class in the image; and

training a classifier in a machine learning manner, for the target object of an unknown class comprised in the automatic clustering result, in order to recognize the target object of the unknown class in the image.

2. The method for recognizing the target object in the image according to claim 1 , wherein training the classifier in the machine learning manner, for the target object of the unknown class comprised in the automatic clustering result, in order to recognize the target object of the unknown class in the image comprises:

performing classification tagging for the unknown class comprised in the automatic clustering result; and

training the classifier for the unknown class with a classification tag using a transfer learning method, in order to recognize the unknown target object in the image.

3. The method for recognizing the target object in the image according to claim 2 , wherein performing classification tagging for the unknown class comprised in the automatic clustering result comprises acquiring, in a human computer interaction manner, classification tag information entered by a user.

4. The method for recognizing the target object in the image according to claim 2 , wherein performing classification tagging for the unknown class comprised in the automatic clustering result comprises:

searching the Internet for another image whose similarity to the unknown class meets a specified requirement; and

performing classification tagging for the unknown class according to tag information of the image on the Internet.

5. The method for recognizing the target object in the image according to claim 2 , wherein training the classifier for the unknown class with the classification tag using the transfer learning method, in order to recognize the unknown target object in the image comprises:

updating an existing classifier according to the automatic clustering result; and

training the classifier for the unknown class with the classification tag using the transfer learning method and the updated existing classifier, in order to recognize the unknown target object in the image.

6. The method for recognizing the target object in the image according to claim 1 , wherein after performing automatic clustering for features in the image according to the feature data in the uniform expression and the historical clustering result, the method further comprises updating the historical clustering result according to the automatic clustering result.

7. An image processing apparatus, comprising:

a memory; and

a processor coupled to the memory, wherein the processor is configured to:

extract feature data from an image;

transform the extracted feature data into a uniform expression;

perform automatic clustering for features in the image according to the feature data in the uniform expression and a historical clustering result;

group a target object of a known class comprised in the automatic clustering result into the corresponding known class, in order to recognize the target object of the known class in the image; and

train a classifier in a machine learning manner, for a target object of an unknown class comprised in the automatic clustering result, in order to recognize the target object of the unknown class in the image.

8. The image processing apparatus according to claim 7 , wherein the processor is further configured to:

perform classification tagging for the unknown class comprised in the automatic clustering result; and

train the classifier for the unknown class with a classification tag using a transfer learning method, in order to recognize the unknown target object in the image.

9. The image processing apparatus according to claim 8 , wherein the processor is further configured to acquire, in a human computer interaction manner, classification tag information entered by a user.

10. The image processing apparatus according to claim 8 , wherein the processor is further configured to:

search the Internet for another image whose similarity to the unknown class meets a specified requirement; and

perform classification tagging for the unknown class according to tag information of the image on the Internet.

11. The image processing apparatus according to claim 8 , wherein the processor is further configured to:

update an existing classifier according to the automatic clustering result; and

train the classifier for the unknown class with the classification tag using the transfer learning method and the updated existing classifier, in order to recognize the unknown target object in the image.

12. The image processing apparatus according to claim 7 , wherein the processor is further configured to update the historical clustering result according to the automatic clustering result.