METHODS AND SYSTEMS FOR DETECTING AN OBJECT BORDERLINE

• US 20140211023A1
• Filed: 01/30/2013
• Published: 07/31/2014
• Est. Priority Date: 01/30/2013
• Status: Active Grant

• Current Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Original Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Inventors: Fan, Zhigang, Ding, Hengzhou, Zhao, Yonghui, Wang, Shen-Ge
• Granted Patent: US 9,218,538 B2
• Time in Patent Office: Days
• US Class Current: 348/207.1
• CPC Class Codes: G06T 2207/10152 Varying illumination
• CPC Class Codes: G06T 2207/20192 Edge enhancement; Edge preservation
• CPC Class Codes: G06T 7/13 Edge detection
• CPC Class Codes: G06T 7/174 involving the use of two or more images
• CPC Class Codes: G06V 10/141 Control of illumination
• CPC Class Codes: G06V 10/25 Determination of region of interest [ROI] or a volume of interest [VOI]
• CPC Class Codes: G06V 10/48 by mapping characteristic values of the pattern into a parameter space, e.g. Hough transformation

Abstract

Methods and systems for detecting an object borderline. A first image with respect to the object can be captured by an image-capturing unit without a flash light and borderlines of the object can be detected. If the detection is successful, the detected borderlines can be outputted. Otherwise, a second image with respect to the object can be captured by the image-capturing unit by applying a flash light and the borderlines can be detected in the image. A geometric transformation between the two images can then be estimated. Finally, the border lines in the first image can be determined by transforming the borderlines detected in the second image. Such an approach effectively detects the appliance borderlines and avoids artifacts caused by applying flash.

Claims

1. A method for detecting an object borderline, said method comprising:

capturing a first image with respect to an object via an image-capturing unit without a flash light to thereafter detect and output at least one borderline with respect to said object if a detection thereof is successful;

capturing a second image with respect to said object by said image-capturing unit by applying of a flash light to thereafter detect at least one borderline in said second image if a detection is not successful; and

estimating a geometric transformation between said first image and said second image to determine said at least one borderline in said first image by transforming said at least one borderline detected in said second image to detect said object borderline.

2. The method of claim 1 further comprising:

linking said at least one borderline detected in said first image and said second image into a straight line utilizing a transformation algorithm; and

selecting at least two pairs of lines that can best construct a rectangle.

3. The method of claim 2 wherein said transformation algorithm comprises a Hough transformation algorithm.

4. The method of claim 1 further comprising capturing said second image immediately after said first image so that said image-capturing unit position and orientation does not change substantially therebetween.

5. The method of claim 1 further comprising modeling a movement by a translation followed by a small rotation and estimating a translation and rotation parameter by matching said first image and said second image.

6. The method of claim 1 further comprising correlating said second image to said first image utilizing said translation and rotation parameter and maximizing said correlation by said translation and rotation parameter.

7. The method of claim 6 wherein said maximization is initially optimized over said translation and then over said rotation.

8. The method of claim 1 further comprising mapping said at least one borderline detected in said second image to said first image utilizing said translation and rotation parameter.

9. A system for detecting an object borderline, said system comprising:

a processor;

a data bus coupled to said processor; and

a computer-usable medium embodying computer program code, said computer-usable medium being coupled to said data bus, said computer program code comprising instructions executable by said processor and configured for;

capturing a first image with respect to an object via an image-capturing unit without a flash light to thereafter detect and output at least one borderline with respect to said object if a detection thereof is successful;

capturing a second image with respect to said object by said image-capturing unit by applying of a flash light to thereafter detect at least one borderline in said second image if a detection is not successful; and

estimating a geometric transformation between said first image and said second image to determine said at least one borderline in said first image by transforming said at least one borderline detected in said second image to detect said object borderline.

10. The system of claim 9 wherein said instructions are further configured for:

linking said at least one borderline detected in said first image and said second image into a straight line utilizing a transformation algorithm; and

selecting at least two pairs of lines that can best construct a rectangle.

11. The system of claim 10 wherein said transformation algorithm comprises a Hough transformation algorithm.

12. The system of claim 9 wherein said instructions are further configured for capturing said second image immediately after said first image so that said image-capturing unit position and orientation does not change substantially in between.

13. The system of claim 9 wherein said instructions are further configured for modeling a movement by a translation followed by a small rotation and estimating a translation and rotation parameter by matching said first image and said second image.

14. The system of claim 9 wherein said instructions are further configured for correlating said second image to said first image utilizing said translation and rotation parameter and maximizing said correlation by said translation and rotation parameter.

15. The system of claim 14 wherein said maximization is initially optimized over said translation and then over said rotation.

16. The system of claim 9 wherein said instructions are further configured for mapping said at least one borderline detected in said second image to said first image utilizing said translation and rotation parameter.

17. A processor-readable medium storing code representing instructions to cause a process for detecting an object borderline, said code comprising code to:

capture a first image with respect to an object via an image-capturing unit without a flash light to thereafter detect and output at least one borderline with respect to said object if a detection thereof is successful;

capture a second image with respect to said object by said image-capturing unit by applying of a flash light to thereafter detect at least one borderline in said second image if a detection is not successful; and

estimate a geometric transformation between said first image and said second image to determine said at least one borderline in said first image by transforming said at least one borderline detected in said second image to detect said object borderline.

18. The processor-readable medium of claim 17 wherein said code further comprises code to:

link said at least one borderline detected in said first image and said second image into a straight line utilizing a transformation algorithm; and

select at least two pairs of lines that can best construct a rectangle.

19. The processor-readable medium of claim 18 wherein said transformation algorithm comprises a Hough transformation algorithm.

20. The processor-readable medium of claim 17 wherein said code further comprises code to:

capture said second image immediately after said first image so that said image-capturing unit position and orientation does not change substantially therebetween; and

model a movement by a translation followed by a small rotation and estimating a translation and rotation parameter by matching said first image and said second image.