how to obtain the threshold 25% of the gray intensities?

Hi, i want to binarize an image by using thresholding.how to obtain threshold 25% of the gray intensities? (25% of the lower gray intensities are discarded) please gudie me.

Answer 1

[Image Analyst]

Try this

binarizedImage = grayImage > 0.25 * intmax(class(grayImage));

That should work for uint8 or uint16 images and produce a logical image of the same size.

[fred bnm]

thanks,but what is intmax and class??

[Image Analyst]

class() tells you what class a variable is, like uint8 or uint16. intmax() gives you the maximum numerical value for the class you pass it, so for uint8 it will return 255 and for uint16 it will return 65,535.

I did that because I wanted a solution that will work regardless of how many bits your image was. If you know what it is, then you can compute it and use it directly. For example

binarizedImage = grayImage > 0.25 * 255; % For a uint8 8 bit image.

or

binarizedImage = grayImage > 63.75;

[fred bnm]

ok.is this code correct if image was in double class ?

    binarizedImage = grayImage > 0.25 * 1

[Image Analyst]

Maybe. Depends on what the range of the image was. This is perhaps more what you want since it's 25% of the max value present in grayImage, regardless of what it's range is:

binarizedImage = grayImage > 0.25 * max(grayImage(:));

Note, that will give you the pixels with the brightest 75% of the intensity values, but this will most likely not mean you have an area fraction of 75% - in case that is what you wanted. In other words, this will not give you an image where 75% of the pixels are white and 25% are black. If you want that, then let me know - it's a different calculation.

[fred bnm]

this is (maximum) More acceptable Compared to the previous.Because that is more adaptable than any image.in fact I am looking for pixels that have 75% of the total gray level image. if you know another method that is suitable for thresholding for Include pixels on edges Please explain.

[Image Analyst]

You have to be more precise in your wording. "pixels that have 75% of the total gray level image" is not precise enough to know what you mean. Do you mean a gray level such that when thresholded at that gray level, 75% of the pixels will be selected? Or do you mean the gray level that is at 25% of the max gray level allowed for that class (uint8 or uint16)? Or do you mean the gray level that is 25% of the max gray level that occurs in the image? Or something else. I don't know.