Dear Veeranjaneyulu Toka,

Use image processing techniques based on Mathematical Morphological Operators.

There are bunch of clustering methods, such as:

color-based, pattern-based, shape-based etc.

Depending on your specific working space you need to select the best approach first. 

You can get an idea by reviewing examples here:

http://www.mathworks.com/discovery/image-segmentation.html?refresh=true

http://www.mathworks.com/videos/color-based-segmentation-with-live-image-acquisition-68771.html

Best, 

Dear Veeranjaneyulu Toka,

Usually, the image is processed with a  of low-pass and high pass filters combination. Then a segmentation technique is used.

For example, a useful sequence might be:

Gaussian filter + gradient magnitude filter +  similarity filter +  region growing technique. If you are interested I can send you sample code in C ++ for this type of application

Hi All, Thank you for all your suggestions.. I am experimenting different approaches, will post you my feedback based on results, accuracy is more important in my app.

@Miguel Vera: plz send me the sample code.

Thanks,

Veeru.