Computer vision system and method employing hierarchical object classification scheme

• US 7,043,075 B2
• Filed: 06/27/2002
• Issued: 05/09/2006
• Est. Priority Date: 09/27/2001
• Status: Active Grant

• Current Assignee: Koninklijke Philips Electronics N.V. (Koninklijke Philips N.V.)
• Original Assignee: Koninklijke Philips Electronics N.V. (Koninklijke Philips N.V.)
• Inventors: Gutta, Srinivas
• Primary Examiner(s): CHAWAN, SHEELA C
• Application Number:
  US10/183,821
• Publication Number:
  US 20030059106A1
• Time in Patent Office: 1,412 Days
• Field of Search: 382/224, 382/159, 382/103, 382/156, 382/158, 382/226, 382/227, 707/1, 707/102, 707/6, 706/31, 706/20, 706/25, 348/169
• US Class Current: 382/158
• CPC Class Codes: G06F 18/2414 Smoothing the distance, e.g. radial basis function networks [RBFN]
• CPC Class Codes: G06F 18/24323 Tree-organised classifiers
• CPC Class Codes: G06T 7/20 Analysis of motion motion estimation for coding, decoding, compressing or decompressing digital video signals H04N19/43, H04N19/51
• CPC Class Codes: G06V 20/52 Surveillance or monitoring of activities, e.g. for recognising suspicious objects recognising microscopic objects G06V20/69
• CPC Class Codes: G08B 13/19602 Image analysis to detect motion of the intruder, e.g. by frame subtraction
• CPC Class Codes: G08B 13/19613 Recognition of a predetermined image pattern or behaviour pattern indicating theft or intrusion
• CPC Class Codes: G08B 21/0423 detecting deviation from an expected pattern of behaviour or schedule
• CPC Class Codes: G08B 21/0476 Cameras to detect unsafe condition, e.g. video cameras

Abstract

A method and apparatus are disclosed for classifying objects using a hierarchical object classification scheme. The hierarchical object classification scheme provides candidate classes with an increasing degree of specificity as the hierarchy is traversed from the root node to the leaf nodes. Each node in the hierarchy has an associated classifier, such as a Radial Basis Function classifier, that determines a probability that an object is a member of the class associated with the node. The nodes of the hierarchical tree are individually trained by any learning technique, such as the exemplary Radial Basis Function Network, that uses appearance-based information of the objects under consideration to classify objects. A disclosed recognition scheme uses a decision criterion based upon recognition error to classify objects.

Claims

1. A method for classifying an object in image data, comprising:

detecting an object in said image data;

classifying said object using a hierarchical object classification scheme defined by providing a candidate class for the detected object with an increasing degree of specificity as a hierarchy of classes is traversed from a root node to a leaf node; and

outputting a class label to identify said candidate class to which the detected object corresponds to and a probability value to indicate a probability with which the detected object belongs to said candidate class.

2. The method of claim 1, wherein said hierarchical object classification scheme employs appearance based feature data.

3. The method of claim 1, wherein said detecting comprises subtracting a background scene from the image data.

4. The method of claim 1, further comprising filtering out unintended moving objects from the detected moving object.

5. The method of claim 4, wherein said filtering comprises filtering out the unintended moving objects according to a detected speed and aspect ratio of the detected moving object.

6. The method of claim 1, further comprising extracting at least two of x-gradient, y-gradient, and combined xy-gradient features from the detected object.

7. The method of claim 1, wherein the classifying comprises using a classifier associated with one or more nodes in said hierarchical object classification scheme.

8. The method of claim 7, wherein said classifier comprises a Radial Basis Function Network for training and classifying the detected object.

9. The method of claim 1, wherein the classifying comprises outputting said class label based upon a recognition error.

10. The method of claim 1, wherein the classifying comprises evaluating a recognition error of nodes in said hierarchical object classification scheme using a top-down approach.

11. A method for classifying an object in image data, comprising:

detecting an object in said image data;

evaluating a recognition error for said object for one or more nodes in a hierarchical object classification scheme, until a node having a lowest recognition error is identified, said hierarchical object classification scheme defined by providing a candidate class for the detected object with an increasing degree of specificity as a hierarchy of classes is traversed from a root node to a leaf node; and

outputting a class label to identify said candidate class to which the detected object corresponds to and a probability value to indicate a probability with which the detected object belongs to said candidate class.

12. The method of claim 11, wherein the evaluating comprises using a classifier associated with one or more nodes in said hierarchical object classification scheme.

13. The method of claim 12, wherein said classifier comprises a Radial Basis Function Network for training and classifying at least one the detected object.

14. The method of claim 11, wherein the evaluating comprises evaluating said recognition error of nodes in said hierarchical object classification scheme using a top-down approach.

15. A system for classifying an object in image data, comprising:

a memory that stores computer-readable code; and

a processor operatively coupled to said memory, said processor configured to implement said computer-readable code, said computer-readable code configured to;

detect an object in said image data;

classify said object using a hierarchical object classification scheme defined by providing a candidate class for the detected object with an increasing degree of specificity as a hierarchy of classes is traversed from a root node to a leaf node; and

output a class label to identify said candidate class to which the detected object corresponds to and a probability value to indicate a probability with which the detected object belongs to said candidate class.

16. The system of claim 15, wherein said hierarchical object classification scheme employs appearance based feature data.

17. The system of claim 15, wherein said processor is further configured to subtract a background scene from the image data.

18. The system of claim 15, wherein said processor is further configured to filter out unintended moving objects from detected moving objects.

19. The system of claim 15, wherein said processor is further configured to use a classifier associated with one or more nodes in said hierarchical object classification scheme.

20. A system for classifying an object in image data, comprising:

a memory that stores computer-readable code; and

a processor operatively coupled to said memory, said processor configured to implement said computer-readable code, said computer-readable code configured to;

detect an object in said image data;

evaluate a recognition error for said object for one or more nodes in a hierarchical object classification scheme, until a node having a lowest recognition error is identified, said hierarchical object classification scheme defined by providing a candidate class for the detected object with an increasing degree of specificity as a hierarchy of classes is traversed from a root node to a leaf node; and

output a class label to identify said candidate class to which the detected object corresponds to and a probability value to indicate a probability with which the detected object belongs to said candidate class.

21. An article of manufacture for classifying an object in image data, comprising:

a computer-readable medium having computer-readable code means embodied thereon, said computer-readable program code means comprising;

detecting an object in said image data;

classifying said object using a hierarchical object classification scheme defined by providing a candidate class for the detected object with an increasing degree of specificity as a hierarchy of classes is traversed from a root node to a leaf node; and

outputting a class label to identify said candidate class to which the detected object corresponds to and a probability value to indicate a probability with which the detected object belongs to said candidate class.

22. An article of manufacture for classifying an object in image data, comprising:

a computer-readable medium having computer-readable code means embodied thereon, said computer-readable program code means comprising;

detecting an object in said image data;

evaluating a recognition error for said object for one or more nodes in a hierarchical object classification scheme, until a node having a lowest recognition error is identified, said hierarchical object classification scheme defined by providing a candidate class for the detected object with an increasing degree of specificity as a hierarchy of classes is traversed from a root node to a leaf node; and

outputting a class label to identify said candidate class to which the detected object corresponds to and a probability value to indicate a probability with which the detected object belongs to said candidate class.