If I understand you. You can do this easily with any image processing package.

Simply create your binary image, invert if necessary, so that the selected area is black and the remaining area is white.

Then using a simpe image arithmetic tool (again this is available in nearly every processing package) subtract the binary image from the original grey scale image. The area outside your selected region will be black and the rest of your image will have the grey scale output.

If you don't have one available you could write something to do this.

Dear

you can do that by applying  arithmetic  operations and thresholding via the matlab package

Regards

Dear Mahesh

You can do that by specifying two threshold values upper and lower, and then create two new bitmaps (foreground, and background) , if the pixel value lies between the two threshold values you can set that pixel in the foreground image, else draw the pixel value in the background.

hope that would be useful,

Best,

By using threshold value

Using MeVisLab you can achieve this with the "IntervalThreshold" module.

Snake segmentation could be a possible solution.

Hi,

The following ImageJ macro will do what you need I think:

open();

rename("input");

run("Duplicate...", "title=mask");

setThreshold(100, 200);

setOption("BlackBackground", false);

run("Convert to Mask");

imageCalculator("Min create", "input","mask");

selectWindow("Result of input");

rename("output");

It assumes a greyscale image as input. It crerates a copy of this for the purpose of making a masking image.The is thresholded with a value range of 100 and 200 (as defined in the "setThreshold" command. You can change this to whatever you want.  The thresholded image is converted to a mask. "Convert to Mask sets all pixels that are within your range  to 255 and the remaining ones  to 0. The imagecalculator  then creates your output image using the "Minimum" function where the values within range are kept and the masked pixels are set to 0.

ImageJ is freely available from imagej.nih.gov/ij/.

Go to Plugins>New>macro and copy and past my macrotext in there and hit the "run" button. The filechooser dialog should open, select your file, and after a little bit your result is in the window called "Output".

hope that helps,

Rob

. 

As some of friends mentioned above, you can use the histogram-based technique. So you need to plot your image histogram then find the Grey matter threshold and apply to your image. 

This method is not 100% accurate but it's fast and pretty well.

Hope it helps,

Saman

Use thresholding in MATLAB for a given range of grey level intensity -- this needs to be selected manually either from a histogram or by trial and error