Object detection

• US 7,421,149 B2
• Filed: 12/08/2004
• Issued: 09/02/2008
• Est. Priority Date: 12/11/2003
• Status: Expired due to Fees

• Current Assignee: Sony United Kingdom Limited (Sony Group Corp.)
• Original Assignee: Sony United Kingdom Limited (Sony Group Corp.)
• Inventors: Haynes, Simon Dominic, Living, Jonathan, Porter, Robert Mark Stefan, Rambaruth, Ratna
• Primary Examiner(s): COUSO, YON JUNG
• Application Number:
  US11/007,820
• Publication Number:
  US 20050129276A1
• Time in Patent Office: 1,364 Days
• Field of Search: 382/100, 382/118, 382/224, 382/228, 382/286, 382/291, 382/103, 382/282, 382/168, 382/171, 382/172, 382/155, 382/160
• US Class Current: 382/291
• 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 30/2504 Coarse or fine approaches, e.g. resolution of ambiguities or multiscale approaches
• CPC Class Codes: G06V 40/161 Detection; Localisation; Normalisation

Abstract

A system and associated methodology of detecting objects using multiple sets of object probability data relating to different respective expected object sizes relative to the image size. A list of locations in an image where an object has been detected is stored, the list defining a probability data set and a probability value for each detected object. Object locations in a set of object probability data as locations at which a local maximum probability value exceeds a threshold probability value are detected and, if that location is substantially similar to a location previously stored in the list, where the local maximum probability value exceeds that relating to the previously stored location. The storing is repeated successively in respect of each set of probability data.

Claims

1. A method of detecting objects using multiple sets of object probability data relating to different respective expected object sizes relative to the image size, said method comprising:

storing a list of locations in an image where an object has been detected, said list defining a probability data set and a probability value for each detected object;

detecting object locations in a set of object probability data as locations at which a local maximum probability value exceeds a threshold probability value and, if that location is substantially similar to a location previously stored in said list, where the local maximum probability value exceeds that relating to the previously stored location; and

repeating step the storing successively in respect of each set of probability data.

2. The method according to claim 1, in which step the storing includes storing data representing an image area relating to each detected object.

3. The method according to claim 2, in which, in the detecting, a local maximum is detected to be substantially similar to a stored object location if said local maximum lies within the image area relating to said stored object.

4. The method according to claim 1, in which the detecting further comprises:

detecting a maximum probability within a current probability data set;

altering at least said maximum probability to equal said threshold value, if the detected maximum probability exceeds the threshold probability value; and

repeating steps the detecting and altering until said maximum probability within the current set does not exceed said threshold value.

5. The method according to claim 4, in which the altering includes altering said maximum probability and probability values within an image area around that maximum to equal said threshold value.

6. The method according to claim 5, in which the image area around said maximum corresponds to the expected object size used in the generation of that set of probability data.

7. The method according to claim 1, further comprising:

generating a set of probability data by combining test blocks of said image with blocks representative of object features.

8. The method according to claim 7, further comprising:

scaling said image and/or said test blocks and/or said blocks indicative of object features before their combination to generate said multiple sets of probability data.

9. The method according to claim 1, further comprising:

generating each set of probability data once the detecting has been applied to the preceding set.

10. The method according to claim 1, in which said objects are faces.

11. A computer readable storage medium including encoded computer program instructions which cause a computer to implement a method according to claim 1.

12. An object detection apparatus for detecting objects using multiple sets of object probability data relating to different respective expected object sizes relative to the image size, said apparatus comprising:

a storage configured to store a list of locations in an image where an object has been detected, said list defining a probability data set and a probability value for each detected object; and

logic, operable successively in respect of each set of probability data, to detect object locations in a set of object probability data as locations at which a local maximum probability value exceeds a threshold probability value and, if that location is substantially similar to a location previously stored in said list, where said local maximum probability value exceeds that relating to the previously stored location.

13. The apparatus according to claim 12, in which said storage is arranged to store data representing an image area relating to each detected object.

14. The apparatus according to claim 13, in which a local maximum is detected to be substantially similar to a stored object location if said local maximum lies within the image area relating to the stored object location.

15. The apparatus according to claim 12, further comprising logic configured to generate a set of probability data by combining test blocks of said image with blocks representative of object features.

16. The apparatus according to claim 15, further comprising logic configured to scale said image and/or said test blocks and/or said blocks indicative of object features before their combination to generate said multiple sets of probability data.

17. The apparatus according to claim 12, in which said objects are faces.

18. A video conferencing apparatus comprising the object detection apparatus according claim 12.

19. A surveillance apparatus comprising the object detection apparatus according claim 12.

20. A camera apparatus comprising the object detection apparatus according to claim 12.