The problem about the display of image(Matlab2014b)

Hello, everyone, the first part is the demo from help.

    I=fitsread('solarspectra.fts');
    I=mat2gray(I);
    imshow(I);

What troubles me a lot is that if I leave out the second sentence %I=mat2gray; the image will be blank.Could U tell me the reason for that? Maybe the format is .mat? I am not sure. Thanks in advance. Hang Yang

[KALYAN ACHARJYA]

fts >> Flexible Image Transport System >> major feature having store non-image data>>image metadata is stored in a human-readable ASCII header.

So you have to convert the metadata matrix to intensity image using mat2gray function.

For better clarification just take any matrix and show as the image Again perform the imshow on the same matrix after converting to mat2gray. You will see the difference.

[Image Analyst]

Kalyan, can you please put answers in the Answers section below (like I did)? You can even get credit for your answers if you put them in the right place.

[Y]

Thanks a lot. Maybe I understand what you said before. Take any matrix as an example.

if true
 b=[1,2;3,4];
 subplot(1,2,1);
 imshow(b);
 subplot(1,2,2);
 c=mat2gray(b);
 imshow(c);
end
I can see difference between two picture. 
The first one is blank due to the fact that we could not see
metadata directly and we need to convert it.
Is it right.
Best regards.

[Guillaume]

The issue, as explained by the two actual answers, has nothing to do with metadata (whatever is meant by that in this case) and only to do with the range of values in the matrix.

Answer 1

[Image Analyst]

If you have a floating point image with values below 0, they will show up as black.

If you have a floating point image with values above 1, they will show up as white.

To get around this use [] in imshow():

imshow(I, []);

[Y]

Thanks all the same.

Answer 2

[Guillaume]

The image is not blank. It is all white.

Matlab assumes that an image of type double is in the range [0-1]. 0 being black, 1 being white. Any value below 0 is considered to be 0 (black), any value above 1 is considered to be 1 (white). This is most likely the case with your matrix where most (all) values are > 1.

As per its documentation mat2gray rescale your matrix to the [0-1] range, so it conforms to the above.

If you don't want to use mat2gray another option is to tell matlab to use the full range of intensities in your matrix. As per the documentation of imshow:

imshow(I, [])  %will use the full image range, whatever that is, instead of [0-1] for double

[Y]

Hh,thanks a lot.I start to learn image processing with Matlab. You are kindness.

[Image Analyst]

If you're just starting, make sure you check out my tutorials, especially the Image Segmentation Tutorial http://www.mathworks.com/matlabcentral/fileexchange/?term=authorid%3A31862&sort=downloads_desc