How to get diminished image?

This is the link for my input image.. http://tinypic.com/r/4h60dv/6 What technique is involved behind to look like the output image.... http://tinypic.com/r/nx3kvp/6

[Walter Roberson]

Is this intended to be a question along the same lines as http://www.mathworks.co.uk/matlabcentral/answers/47786-how-to-make-a-image-to-look-very-diminish ? The second image you give here does not appear to be any lower resolution than the first.

[Image Analyst]

The technique seems like it's "image segmentation" - is that what you were looking for?

di·min·ish    verb (used with object) 1. to make or cause to seem smaller, less, less important, etc.; lessen; reduce. 2. Architecture . to give (a column) a form tapering inward from bottom to top. 3. Music . to make (an interval) smaller by a chromatic half step than the corresponding perfect or minor interval. 4. to detract from the authority, honor, stature, or reputation of; disparage.

Not sure how diminish applies here.

[Walter Roberson]

I don't even see any segmentation, not unless the input and outputs are reversed.

[Image Analyst]

My guess is that he incorrectly reversed the images, as you say. Actually they're not even the same image where the desired organ is outlined in green. The output image looks like some kind of superposition of a bunch of slices, or a partial reconstruction from a very limited number of projections. The two images look only vaguely related. He should take more care in selecting which images to upload, and in describing them.

[Sivakumaran Chandrasekaran]

Thanks Image Analyst and Walter. But, my question is correct. For more details,, you can refer the paper titled "Medical Image Segmentation by Combining Graph Cuts and Oriented Active Appearance Models"

[Image Analyst]

Then we can't help you because you haven't accurately specified what you mean by "technique" when you say "What technique is involved behind to look like the output image". So the best I can do is to say that the technique is CT or maybe the technique is filtered back projection (which is the reconstruction technique for CT imagery).

Or, like I already said the technique is segmentation (but that assumed you reversed your input and output images), or, more accurately, the technique is "Medical Image Segmentation by Combining Graph Cuts and Oriented Active Appearance Models" - but then you already know that so now please explain why you asked.

But we still doubt that your input image is really the input image for the image you claim is your output image (take another look at them please). I don't think you took your single input image, and sent it into some function, and your output image popped out, but whatever......good luck with it.

[Sivakumaran Chandrasekaran]

Thanks Image Analyst.. I shall work it on and will ask if i get any doubt.