Method and a device for detecting graphic symbols

• US 20080253656A1
• Filed: 04/12/2007
• Published: 10/16/2008
• Est. Priority Date: 04/12/2007
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Shenhav, Tom, Schwartzberg, Yoav
• Granted Patent: US 8,023,725 B2
• Time in Patent Office: Days
• US Class Current: 382/181
• CPC Class Codes: G06T 2207/30244 Camera pose
• CPC Class Codes: G06T 7/73 using feature-based methods
• CPC Class Codes: G06V 20/63 Scene text, e.g. street names
• CPC Class Codes: G06V 30/10 Character recognition

Abstract

A method for identifying a predefined graphic symbol having a plurality of graphical characters. The method comprises the following steps: a) receiving a digital image having a plurality of pixels depicting a scene, b) identifying a plurality of first groups of contiguous pixels in the proximity of one another, members of each one the first group having a first common pixel defining property, and c) identifying at least one of the plurality of first groups as one of the plurality of graphical characters, thereby detecting the predefined graphic symbol in the digital image.

Claims

1. A method for identifying a predefined graphic symbol having a plurality of graphical characters, said method comprising:

a) receiving a digital image having a plurality of pixels depicting a scene;

b) identifying a plurality of first groups of contiguous pixels in the proximity of one another, members of each one said first group having a first common pixel defining property; and

c) identifying at least one of said plurality of first groups as one of the plurality of graphical characters, thereby detecting the predefined graphic symbol in said digital image.

2. The method of claim 1, further comprising a step a1) between step a) and b) of identifying a second group of contiguous pixels, said second group having a second common pixel defining property, wherein said plurality of first groups is in the proximity of said second group.

3. The method of claim 1, further comprising a step b1) between step b) and c) of identifying an order of said plurality of second groups in relation to the position of each one of them in said digital image, said identifying of step c) being performed according to the place of said at least one of said plurality of first groups in said order.

4. The method of claim 1, wherein said identifying of step b) comprises a step of substantially enclosing each said first group in a respective first bounding-box, said first bounding-boxes being used during step c).

5. The method of claim 2, wherein said step a1) identifying of step b) comprises a step of substantially enclosing said plurality of first groups in a second bounding-box, said second bounding-box being used during step c).

6. The method of claim 3, further comprising a step between step b) and b1) of identifying a line that approximately intersects the geometric center of each said first group, said order being determined according to relative position each said each said first group along said line.

7. The method of claim 6, wherein said line is a regression line.

8. The method of claim 3, further comprising a step between step b) and b1) of detecting and removing outlier groups from said plurality of first groups.

9. The method of claim 1, further comprising the following steps:

d) identifying a homography matrix according to at least one of said identified plurality of first groups; and

e) using said homography matrix for augmenting a virtual object with said digital image.

10. The method of claim 9, wherein said homography matrix is calculated according to at least four reference points of said identified plurality of first groups.

11. The method of claim 10, each three of said at least four references are non-collinear.

12. The method of claim 5, further comprising a step between step a) and b) of generating a binary map image based on said second bounding-box, said identifying of step b) being performed according to said binary map.

13. The method of claim 12, wherein said generating comprises a step of applying a color property threshold on the color property of the plurality of pixels.

14. The method of claim 13, wherein said color property threshold is a member of the following group:

a constant threshold, a dynamic threshold, and a Niblack threshold.

15. The method of claim 1, wherein said identifying of step b) comprises a step of finding connected components, each said connected component being tagged as one of said plurality of first groups.

16. The method of claim 1, wherein said identifying of step b) comprising a step of removing members of said plurality of first groups that does not comply with a set of predefined geometric characteristics.

17. The method of claim 16, wherein said set of predefined geometric characteristics comprises a member of the following group:

a predefined aspect ratio between the width and the height, having an area in a predefined range, having an area in a predefined range in relation to an area that encloses said first group, having an area in a predefined range in relation at least a portion of said plurality of second bounding-boxes.

18. A method for augmenting a virtual object with a digital image, said method comprising:

a) receiving the digital image, said digital image depicts a graphic symbol;

b) identifying said graphic symbol in said digital image; and

c) augmenting the digital image with the virtual object, to create an augmented-reality image for presentation to a user.

19. The method of claim 18, wherein said virtual object comprises a graphical object, said augmented-reality image having said graphical object positioned in close proximity to said graphic symbol.

20. The method of claim 18, wherein the virtual object, the digital image, and said augmented-reality image are video images.

21. The method of claim 18, further comprising a step between steps a) and b) of identifying a homography matrix according characteristics of said graphic symbol, said augmenting comprises a step of using said homography matrix.

22. The method of claim 21, wherein said homography matrix is calculated according to at least four reference points identified in said graphic symbol.

23. The method of claim 21, wherein said augmented-reality image is based on the digital image.

24. An apparatus suitable for viewing a digital image depicting a scene, the apparatus comprises:

an image input device having a connection to an image sensor, said image input device adapted to receive the digital image from said image sensor;

a graphic symbol detection module adapted to detecting a label in the scene;

an image augmenting module adapted to augment the scene with a virtual object, to create an augmented-reality image; and

a display device for displaying said augmented-reality image.

25. The apparatus of claim 24, wherein said augmented-reality image is based on the digital image.