Image binarization based on grey membership parameters of pixels

1. A method of binarizing an image, comprising:

partitioning the image into a plurality of image segments, each image segment having a plurality of image pixels;

partitioning each of the plurality of image segments into a plurality of subsegments, each image subsegment having a plurality of image pixels;

estimating a grey membership parameter for each image pixel for each of the plurality of image segments and subsegments;

combining the grey membership parameter for each of the plurality of image pixels from each of the plurality of image segments and subsegments to estimate a net grey membership parameter for each image pixel;

assigning black or white to each of the plurality of image pixels based on the estimated net grey membership parameter of each respective one of the plurality of image pixels;

enhancing a contrast level of each of the plurality of image segments and subsegments to estimate a contrast-stretched grey value for each image pixel of each of the plurality of image segments and sub segments;

accessing a color image or a grey image of an object;

estimating a grey value for each of the plurality of image pixels of the color image or grey image;

estimating mean and standard deviation of grey values of image pixels of each of the plurality of image segments and subsegments;

comparing the standard deviation with a pre-determined threshold; and

assigning a first grey membership parameter to image pixels of the plurality of image segments having the standard deviation below the pre-determined threshold and assigning a second grey membership parameter to image pixels of the plurality of image segments having the standard deviation above the pre-determined threshold.

2. The method of claim 1 , wherein the first grey membership parameter is estimated by dividing the mean of the grey values with a pre-determined maximum grey value of the image pixel.

3. The method of claim 1 , wherein the second grey membership parameter is estimated by dividing the contrast-stretched grey value with a pre-determined maximum grey value of the image pixel.

4. The method of claim 1 , wherein the enhancing the contrast level comprises stretching the contrast level of each of the plurality of image pixels using histogram equalization.

5. The method of claim 4 , wherein a number of grey levels for stretching the contrast level comprises 256 and the maximum grey value of an image pixel is comprises 255.

6. The method of claim 1 , wherein the combining the grey membership parameter comprises determining a weighted average of grey membership parameters for each of the plurality of image pixels from each of the plurality of image segments and subsegments to estimate the net grey membership parameter.

7. The method of claim 6 , wherein the weighted average comprises a function of a number of the plurality of image segments or subsegments within each image segment.

8. The method of claim 6 , wherein the weighted average is a function of a square of a number of the plurality of image segments or subsegments within each image segment.

9. The method of claim 1 , wherein the assigning black or white color comprises: comparing the net grey membership parameter with the pre-determined threshold; and

assigning black color to the image pixels having the net grey membership parameter below the pre-determined threshold and assigning white color to the image pixels having the net grey membership parameter above the pre-determined threshold.

10. A method of binarizing an image, comprising:

partitioning the image into a plurality of image segments, each image segment having a plurality of image pixels;

partitioning each of the plurality of image segments into subsegments, each subsegment having a plurality of image pixels;

enhancing a contrast level of each of the plurality of image segments and subsegments to estimate a contrast-stretched grey value for each image pixel of each of the plurality of image segments and subsegments;

estimating a grey membership parameter for each image pixel for each of the plurality of image segments and subsegments based upon the contrast-stretched grey value of the image pixel;

determining a weighted average of grey membership parameters for each of the plurality of image pixels from each of the plurality of image segments and subsegments to estimate a net grey membership parameter for each image pixel, wherein the weighted average is a linear function of a number of the plurality of image segments or subsegments within each image segment; and

assigning black or white to each of the plurality of image pixels based on the estimated net grey membership parameter of each respective one of the plurality of image pixels, wherein k is a number of partitioning levels, the grey membership parameter for a pixel p(x,y) is membership [y][x][k], and the net grey membership parameter net membership[y][x] is estimated in accordance with a following relationship:

net_membership

⁡

[

y

]

⁡

[

x

]

=

∑

k

=

startlevel

endlevel

⁢

⁢

membership

⁡

[

y

]

⁡

[

x

]

⁡

[

k

]

×

(

k

+

1

)

∑

k

=

startlevel

endlevel

⁢

⁢

(

k

+

1

)

.

11. The method of claim 10 , comprising assigning black color to the image pixels having the net grey membership parameter below 0.5 and assigning white color to the image pixels having the net grey membership parameter above 0.5.

12. An image binarization system, comprising:

a memory circuit configured to store an input image; and

an image processing circuit configured to:

partition the input image into a plurality of image segments and subsegments, each image segment and subsegment having a plurality of image pixels,

estimate a net grey membership parameter for each image pixel of each of the plurality of image segments and subsegments by combining individual grey membership parameters of each image pixel from each of the plurality of image segments and subsegments,

enhance a contrast level of each of the plurality of image segments and subsegments using histogram equalization,

estimate the individual grey membership parameter of each image pixel based upon a contrast-stretched grey value for the image pixel,

access a color image or a grey image of an object,

estimate a grey value for each of the plurality of image pixels of the color image or grey image,

estimate mean and standard deviation of the grey values of image pixels of each of the plurality of image segments and subsegments,

compare the standard deviation with a pre-determined threshold, and

assign a first grey membership parameter to image pixels of the plurality of image segments having the standard deviation below the pre-determined threshold and assigning a second grey membership parameter to image pixels of the plurality of image segments having the standard deviation above the pre-determined threshold.

13. The image binarization system of claim 12 , wherein the image processing circuit is further configured to determine a weighted average of grey membership parameters for each of the plurality of image pixels from each of the plurality of segments and subsegments to estimate the net grey membership parameter.

14. The image binarization system of claim 12 , wherein the image processing circuit is further configured to generate a binary output image by assigning black or white color to each image pixel based upon the estimated net grey membership parameter of the respective pixel.

15. The image binarization system of claim 14 , wherein the image processing circuit is further configured to substantially reduce noise in the binary output image using morphological open-close technique.