Method and system for binarizing an image

1. A method for binarizing an image, the method comprising:

acquiring a copy of the image in an electronic memory in an input device;

establishing boundaries of image objects of the image based on contrast values for the image objects;

classifying each image object as either suspect or non-suspect;

creating a local binarization threshold map, the local binarization threshold map comprising threshold binarization values associated with image objects classified as non-suspect; and

creating a global binarization threshold map for the entire image by expanding the local binarization threshold map to cover the entire image.

2. The method of claim 1 , the method further comprising:

creating a binarized image based on the global binarization threshold map.

3. The method of claim 2 , the method further comprising:

applying an optical character recognition (OCR) algorithm to the binarized image.

4. The method of claim 1 , wherein establishing the boundaries of the image objects is based on contrast values for the image objects.

5. The method of claim 1 , further comprising smoothing the image prior to establishing the boundaries.

6. The method of claim 1 , wherein classifying an image object comprises specifying a penalty and classifying the image object as suspect when it exceeds the penalty and as non-suspect when it does not exceed the penalty.

7. The method of claim 6 , wherein the penalty comprises a criterion selected from the group consisting of a number of pixels for the image object, an area and linear dimensions of the image object, a number of inside objects, and a ratio of the sides of the image object.

8. The method of claim 6 , wherein the classifying is performed locally based on segments of the image and a mean penalty for each segment.

9. The method of claim 8 , wherein the classifying comprises deeming an image object suspect in all segments when the object is suspect in at least one segment.

10. The method of claim 1 , wherein expanding the local binarization threshold map is based on an approximation technique to approximate threshold binarization values for pixels not in any local binarization threshold map.

11. A computer-readable medium embodying a set of instructions which, when executed by a computer, causes the computer to perform a method for binarizing an image which method comprises:

acquiring a copy of the image in an electronic memory of the computer;

establishing boundaries of image objects of the image;

classifying each image object as either suspect or non-suspect;

creating a local binarization threshold map comprising threshold binarization values associated with image objects classified as non-suspect; and

expanding the local binarization threshold map to cover the entire image thereby

creating a global binarization threshold map for the entire image.

12. The computer-readable medium of claim 11 , the instructions further being operable to create a binarized image based on the global binarization threshold map.

13. The computer-readable medium of claim 12 , the instructions further comprising an optical character recognition (OCR) algorithm operable to be applied to the binarized image.

14. The computer-readable medium of claim 11 , wherein establishing the boundaries of the image objects is based on contrast values for the image objects.

15. The computer-readable medium of claim 11 , further being operable to smooth the image prior to establishing the boundaries.

16. The computer-readable medium of claim 11 wherein classifying an image object comprises specifying a penalty and classifying the image object as suspect when the image object exceeds the penalty and as non-suspect when the image object does not exceed the penalty.

17. The computer-readable medium of claim 11 wherein expanding the local binarization threshold map is based on an approximation technique to approximate threshold binarization values for pixels not in any of local binarization threshold map.

18. A system for binarizing an image comprising:

a processor; and

a memory coupled to the processor, the memory storing instructions which when executed by the processor cause the system to perform a method for binarizing an image, the method comprising:

establishing boundaries of image objects of the image based at least in part on contrast values for the image objects;

classifying each image object as either suspect or non-suspect;

creating a local binarization threshold map comprising threshold binarization values associated with image objects classified as non-suspect; and

expanding the local binarization threshold map to cover the entire image thereby creating a global binarization threshold map.

19. The system of claim 18 , wherein classifying an image object comprises specifying a penalty and classifying the image object as suspect when the image object exceeds the penalty and as non-suspect when the image object does not exceed the penalty.