How to apply moment preserving to grayscale document image

Hi how to apply moment preserving to grayscale document image to obtain gray values. how to proceed i need a example code.

[goldensona]

http://mirror.transact.net.au/pub/sourceforge/s/project/sv/svenlib/src/sven/segmentation/Thresholders.java i have found this java code but how to assing this histo value it give me error double[] histo = new double[MAX_VALUE + 1];

[Jing]

What do you mean by 'moment preserving'?

[goldensona]

by applying moment preserving prinicpal for get thresholding to binarize image, like default method for obtain threshold

[Image Analyst]

Nothing is going to happen with your question unless you formulate a proper question. Until then, no one has any clue as to what MATLAB code of yours needs fixing. Please read this tutorial: http://www.mathworks.com/matlabcentral/answers/6200-tutorial-how-to-ask-a-question-on-answers-and-get-a-fast-answer. Also, read this FAQ: http://matlab.wikia.com/wiki/FAQ#Can_you_program_up_the_algorithm_in_this_article_for_me_and_explain_it_to_me.3F

[goldensona]

moment preserving is one of the principal,by applying this principal getting max grayvalues and min gray value from a image. if use normal max and min function it always returns 0,255. for all images . can anyone know about the moment preserving principal and how to apply it in image. if my question is not clear , i will explain again

[Image Analyst]

It is not clear. Please explain it to someone else and have them write it down and paste their explanation back here. Otherwise all I can say is try this:

outputImage = inputImage;

Everything from the input image will be preserved, including ALL moments.

[goldensona]

I want to Convert of Input image to Binary form - By Apply moment-preserving threshold (W. H. Tsai, “Moment-Preserving Thresholding:) to input image to obtain two representative gray values g1 and g2, i have to compute threshold value from it , T = (g1 + g2)/2; using the threshold i need to convert it into binary.

Answer 1

[Image Analyst]

Ah, finally now we're getting somewhere. Moment preserving THRESHOLDING. That thresholding word is important. You really didn't explain the algorithm though. So I have no idea what g1 and g2 are, given some arbitrary gray scale image. And the FAQ applies here.

But let's assume you have them and you calculated a threshold value of T according to your formula. Then you could binarize (threshold) the image into two levels (foreground and background) simply by doing this:

binaryImage = grayImage < T;  % Or you can use <=, >, or >= if you want.

This will not preserve ALL moments, but may preserve one or a few of the moments if T was chosen appropriately.

[goldensona]

g1,g2 is grey scale values from a image.

[Image Analyst]

Yes, I know that. How about 42 and 213 - will those do? If not, then why not?