I don't know which type of image is?

I have few foot ulcer images in that I can't detect the ROI.And the output of entropy and variance is also zero.So can anyone tell me which type of image it is?

Answer 1

[Image Analyst]

Without seeing the image (which people usually do when asking for image processing advice), I'd say that if the variance is zero, you have a completely uniform image.

[riddhi desai]

[riddhi desai]

this kind of image sir

[Image Analyst]

Tell me why that image has a variance of zero! It certainly should NOT.

[riddhi desai]

because i have to find the variance of ROI and my ROI image is black (not able to gert ROI). So variance is zero so tell me the type of this image please m not able to find out.

[Walter Roberson]

The type of image is "an image which the wrong ROI has been used for".

[riddhi desai]

but the wrong roi i don't get the ROI in that image i only gate black image

[Walter Roberson]

Is the all-black ROI the right ROI? I don't think so.

At the very least you should do skin detection; see http://www.mathworks.com/matlabcentral/answers/49435-i-am-basically-trying-to-find-the-skin-in-the-image-so-i-wrote-the-piece-of-code-i-converted-the-im#answer_60411

I have no idea how you hope to get any useful ulcer information out of an image that is only 56 by 132 pixels and has been heavily compressed with lossy JPEG.

[riddhi desai]

I dont get ROI sir. And i know its wrong ROI Sir but it have some specific name or what?

[Walter Roberson]

It has the specific name of "an image which the wrong ROI has been used for". We could name those after you if you need a shorter name. They are "Riddhi Images".

[Image Analyst]

Show the code that you are using to define an ROI from that image. Otherwise there's no way I can tell why you're finding all black.

[riddhi desai]

i am using opening and closing for finding ROI sir

[riddhi desai]

@walter Roberson ,LOL sir i want a technical name .

[Image Analyst]

Ridding, did you read my question? We need the code - the actual lines of code.

[riddhi desai]

fs1=im2bw(fs);
figure,imshow(fs1);title('Binarized Image');
se2=strel('disk',3);
fs2=imclose(fs1,se2);title('Filling of Small Holes');
figure,imshow(fs2);
se1=strel('disk',5);title('Removal of small regions');
fs3=imopen(fs2,se1);
%figure,imshow(fs3);title('Final Processed Image');
for i=1:size(fs,1)
    for j=1:size(fs,2)
        if(fs3(i,j)==1)
            fs4(i,j)=fs(i,j);
            fk1(i,j)=fh(i,j);
            fk2(i,j)=fv(i,j);
        else
            fs4(i,j)=0;
            fk1(i,j)=0;
            fk2(i,j)=0;

          end
      end
  end
  figure,imshow(fs4);title('Extracted Region');

Labelling and Connected Component Analysis %%%%%

[B,L]=bwboundaries(fs3,'noholes');
hold on
for k = 1:length(B)
    boundary = B{k};
    plot(boundary(:,2), boundary(:,1), 'y', 'LineWidth', 4)
end
fk3=cat(3,fk1,fs4,fk2);
fk4=hsv2rgb(fk3);

figure,imshow(fk4);title('Extracted Region as in Original Color Image');

[Walter Roberson]

Your question is like asking what the technical name is for the kind of floor that didn't get swept because you brought the wrong size of broom. Hereafter they will be known as "Riddhi floors", and the adjective "Riddhi" shall apply to anything upon which work did not get done because the wrong tool for the job was brought.

[Image Analyst]

The problem is this line:

fs1=im2bw(fs);

That is far, far from an acceptable method to do segmentation on that type of image. You'll have to come up with a better method.

[riddhi desai]

sir i already apply this code on segmented image.so ,and else i also have other question too can u help me in that also?

[Image Analyst]

Yes, perhaps.

[riddhi desai]

http://in.mathworks.com/matlabcentral/answers/291910-the-for-loop-on-classification-gives-every-time-different-value pls ans of this question