Automatic person meta-data labeller

• US 7,054,481 B2
• Filed: 07/27/2001
• Issued: 05/30/2006
• Est. Priority Date: 08/04/2000
• Status: Expired due to Fees

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: Lloyd-Jones, Daniel John
• Primary Examiner(s): Wu, Jingge
• Assistant Examiner(s): Hung, Yubin
• Application Number:
  US09/915,324
• Publication Number:
  US 20020048397A1
• Time in Patent Office: 1,768 Days
• Field of Search: 382/103, 382115-118, 382/162, 382/164, 382/173, 382/203, 382/165, 382/254, 382/260, 382/298, 382/300, 348/580, 358/451
• US Class Current: 382/162
• CPC Class Codes: G06T 2207/10024 Color image
• CPC Class Codes: G06T 2207/20101 Interactive definition of point of interest, landmark or seed
• CPC Class Codes: G06T 7/0008 checking presence/absence
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06T 7/187 involving region growing; involving region merging; involving connected component labelling
• CPC Class Codes: G06T 7/90 Determination of colour characteristics

Abstract

A method 214 of identifying an object in a target image is disclosed, the method comprising steps of selecting 400 a part of the object having a distinctive color, searching 702 the target image for said part, and tagging 704 the target image if said part is found therein.

Claims

1. A method of identifying, from a set of images, images in which a specified person is present, the method comprising the steps of:

choosing, from the set of images, an indicative image in which the specified person is present;

manually designating an accessory worn by the specified person in the indicative image;

establishing for each image in the set of images a corresponding color segmented image having regions of uniform color, wherein the number of said regions is less than the number of colors in said image;

determining for each image in the set whether the color of the designated accessory in the indicative image matches at least one color region of the corresponding color segmented image; and

tagging the images in the set whose corresponding color segmented images contain a region whose color matches the color of the designated accessory in the indicative image.

2. A method according to claim 1, wherein said method further comprises the step of defining the accessory as an item in a menu prior to said manual designation step, and wherein said manual designation step comprises manually designating the accessory in the menu.

3. A method according to claim 1, wherein:

said manual designation step comprises an additional sub-step of manually designating another accessory worn by the specified person in the indicative image, wherein the other accessory has another color;

said determining step further comprises the step of determining for each image in the set whether the color of the another accessory in the indicative image matches at least another color region of the corresponding color segmented image; and

said tagging step comprises the step of tagging images in the set whose corresponding color segmented images contain both a region whose color matches the color of the designated accessory and another region whose color matches the another color.

4. A method according to claim 1, further comprising, prior to said choosing step, additional steps of:

deriving meta-data for a core set of images;

grouping the core set into one or more event image sets dependent upon the meta-data; and

choosing the set of images from the one or more event image sets.

5. A method according to claim 4, wherein the meta-data comprises time stamps associated with the images in the core set, and said grouping step comprises, in relation to an image in the core set, a sub-step of:

assigning the image to an event image set if an associated time stamp falls within a predetermined event time interval.

6. A method according to claim 4, wherein said meta-data comprises time stamps for the images in the core set, and said grouping step comprises, in relation to an image in the core set of images, a sub-step of:

assigning the image to an event image set if an associated time stamp can be clustered with time stamps associated with other images in the event image set.

7. A method according to claim 1, wherein:

(1) the accessory manually designated in the indicative image, in addition to its color, has at least one of a distinctive size attribute and a distinctive shape attribute;

(2) said determining step comprises at least further sub-steps of;

(a) determining whether a size of a region of the corresponding color segmented image matches the distinctive size; and

(b) determining whether a shape of a region of the corresponding color segmented image matches the distinctive shape; and

(3) said tagging step comprises tagging the images in the set whose corresponding color segmented images contain a region whose color matches the color of the designated accessory in the indicative image, wherein said region also has at least one of;

(a) a shape that matches the shape of the designated accessory in the indicative image; and

(b) a size that matches the size of the designated accessory in the indicative image.

8. An apparatus for identifying, from a set of images, images in which a specified person is present, said apparatus comprising:

means for choosing, from the set of images, an indicative image in which the specified person is present;

means for permitting manual designation of an accessory worn by the specified person in the indicative image;

means for establishing for each image in the set of images a corresponding color segmented image having regions of uniform color, wherein the number of said regions is less than the number of colors in said image;

means for determining for each image in the set whether the color of the designated accessory in the indicative image matches at least one color region of the corresponding color segmented image; and

means for tagging the images in the set whose corresponding color segmented images contain a region whose color matches the color of the designated accessory in the indicative image.

9. An apparatus according to claim 8, further comprising:

means for defining the accessory as an item in a menu; and

means for permitting manual designation of the accessory from the menu.

10. An apparatus according to claim 8, further comprising:

means for deriving meta-data for a core set of images;

means for grouping the core set into one or more event image sets dependent upon the meta-data; and

means for choosing the set of images from the one or more event image sets.

11. An apparatus for identifying, from a set of images, images in which a specified person is present, said apparatus comprising:

a memory configured to store a program; and

a processor configured to execute the program, the program comprising;

(a) code for choosing, from the set of images, an indicative image in which the specified person is present;

(b) code for permitting manual designation of an accessory worn by the specified person in the indicative image;

(c) code for establishing for each image in the set of images a corresponding color segmented image having regions of uniform color, wherein the number of said regions is less than the number of color in said image;

(d) code for determining for each image in the set whether the color of the designated accessory in the indicative image matches at least one color region of the corresponding color segmented image; and

(e) code for tagging the images in the set whose corresponding color segmented images contain a region whose color matches the color of the designated accessory in the indicative image.

12. A computer program embodied in a computer-readable medium, which is configured to instruct a computer to execute a procedure for identifying, from a set of images, images in which a specified person is present, said program comprising:

(a) code for choosing, from the set of images, an indicative image in which the specified person is present;

(b) code permitting manual designation of an accessory worn by the specified person in the indicative image;

(c) code for establishing for each image in the set of images a corresponding color segmented image having regions of uniform color, wherein the number of said regions is less than the number of colors in said image;

(d) code for determining for each image in the set whether the color of the designated accessory in the indicative image matches at least one color region of the corresponding color segmented image; and

(e) code for tagging the images in the set whose corresponding color segmented images contain a region whose color matches the color of the designated accessory in the indicative image.

13. A computer program embodied in a computer-readable medium, according to claim 12, further comprising:

code for defining the accessory as an item in a menu; and

code for permitting manual designation of the accessory from the menu.

14. A computer program embodied in a computer-readable medium, according to claim 12, further comprising:

code for deriving meta-data for a core set of images;

code for grouping the core set into one or more event image sets dependent upon the meta-data; and

code for choosing a desired image set, comprising an indicative image and at least one target image, from the one or more event image sets.

15. A computer program product including a computer readable medium having recorded thereon a computer program which is configured to instruct a computer to execute a procedure for identifying, from a set of images, images in which a specified person is present, the program comprising:

code for choosing, from the set of images, an indicative image in which the specified person is present;

code for permitting manual designation of an accessory worn by the person in the indicative image;

code for establishing for each image in the set of images a corresponding color segmented image having regions of uniform color, wherein the number of said regions is less than the number of colors in said image;

code for determining for each image in the set whether the color of the designated accessory in the indicative image matches at least one color region of the corresponding color segmented image; and

code for tagging the images in the set whose corresponding color segmented images contain a region whose color matches the color of the designated accessory in the indicative image.