Image recognition method

• US 20090028389A1
• Filed: 09/22/2008
• Published: 01/29/2009
• Est. Priority Date: 11/28/2006
• Status: Active Grant

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: Edanami, Ikumi
• Granted Patent: US 8,204,278 B2
• Time in Patent Office: Days
• US Class Current: 382/104
• CPC Class Codes: G06T 2207/30236 Traffic on road, railway or crossing
• CPC Class Codes: G06T 7/68 of symmetry
• CPC Class Codes: G06T 7/73 using feature-based methods
• CPC Class Codes: G06T 7/77 using statistical methods
• CPC Class Codes: G06V 10/42 Global feature extraction by analysis of the whole pattern, e.g. using frequency domain transformations or autocorrelation
• CPC Class Codes: G06V 10/422 for representing the structure of the pattern or shape of an object therefor
• CPC Class Codes: G06V 20/58 Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads

Abstract

According to an aspect of an embodiment, a method for detecting a subject in an image, comprising the steps of: dividing said image into a plurality of regions; calculating a similarity between a feature of one of said regions and the feature of another of said regions; determining a distribution of said similarities corresponding to said regions; and detecting the subject in the image by determining correlation of said distribution with a shape of said subject.

Claims

1. A method for detecting a subject in an image, comprising the steps of:

dividing said image into a plurality of regions;

calculating a similarity between a feature of one of said regions and the feature of another of said regions;

determining a distribution of said similarities corresponding to said regions; and

detecting the subject in the image by determining correlation of said distribution with a shape of said subject.

2. The method according to claim 1, wherein said calculating associates said similarity which is the most similar among the features of another of said regions.

3. The method according to claim 1, further comprising, setting an second region for executing said calculating said similarity on the basis of the position of said image.

4. The method according to claim 3, wherein when said image is projected a three-dimensional space, said calculating sets said second region on the basis of a portion of depth of said three-dimensional space.

5. The method according to claim 1, wherein said subject is a vehicle, said detecting detects said vehicle of said shape of the distribution of the regions having a large value of said difference being distributed on either side of a detection range, said large value of said difference is symmetric, and a small value of said difference being distributed a lower side of said detection range.

6. The method according to claim 1, wherein said subject is a vehicle and said calculating further sets said second region capable of extracting said vehicle on said image.

7. The method according to claim 1, wherein said subject is a vehicle, further comprising the steps of,

determining a position of center of width direction of said vehicle on the basis of symmetry of said distribution of said similarities, and

determining a position of vertical direction in said image by determining correlation of said distribution with a shape of said vehicle.

8. The method according to claim 7, wherein when said image is projected a three-dimensional space, further comprising the step of, calculating a width of said vehicle on the basis of width of the subject in the image and said position of said depth of said three-dimensional space.

9. The method according to claim 8, further comprising the step of, determining a type of said vehicle on the basis of said width of said vehicle.

10. The method according to claim 1, wherein said calculating calculates said similarity after deleting a periodically repeated pattern of a crosswalk in said image.

11. An apparatus for detecting a subject in an image, comprising:

a memory for storing a shape information of said subject; and

a processor for executing a process comprising,

dividing said image into a plurality of regions;

calculating a similarity between a feature of one of said regions and the feature of another of said regions;

determining a distribution of said similarities corresponding to said regions; and

detecting the subject in the image by determining correlation of said distribution with a shape of said subject.

12. A computer readable medium storing a program used in a computer for executing a process for detecting a subject in an image, said process comprising the steps of:

dividing said image into a plurality of regions;

calculating a similarity between a feature of one of said regions and the feature of another of said regions;

determining a distribution of said similarities corresponding to said regions; and

detecting the subject in the image by determining correlation of said distribution with a shape of said subject.