imshow shows image as noisy even though its not

1 次查看（过去 30 天）

PhD Problems 2018-12-16

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/436001-imshow-shows-image-as-noisy-even-though-its-not

评论： Guillaume 2018-12-16

Hi, Im trying to learn a segmentation network using CNN, my network is producing very poor results. I looked at the images and I'm wondering if this is the reason. My input images are stack of images in a .tif file. In Windows image viewer below is what I get:

I am trying to detect and segment the bright spots shown above

But when I open the same in Matlab using imshow() I get

All the information is basically lost. However when I use imagesc() I get the following:

Which is much better, but why are my images not working with my network? Is it reading in the imshow() version output? How do I fix this?

Thanks

edit:

below is the code I am using to read in the tiff files:

fname = 'my_file_with_lots_of_images.tif';
info = imfinfo(fname);
imageStack = [];
numberOfImages = length(info);
for k = 1:numberOfImages
    
currentImage = imread(fname, k, 'Info', info);
    
imageStack(:,:,k) = currentImage;
end 

5 个评论
显示 3更早的评论隐藏 3更早的评论

PhD Problems 2018-12-16

编辑：PhD Problems 2018-12-16

在 MATLAB Online 中打开

imshow(imageStack(:,:,7))
imagesc(imageStack(:,:,7))

My concern is how the data is interpreted by the CNN network, is it how it is shown in imshow()?

Also

imshow(imageStack(:,:,7), [])

produces the following

Thanks

Guillaume 2018-12-16

We need either the raw image (the tif file) or the output of imfinfo to understand what type of image you're dealing with.

Other than a different colour map (grey vs green) your latest display with imshow looks more or less the same as the one you get with the image viewer.

Possibly, the image is indexed, in which case, you're currently not loading the colour map (2nd output of imread). Hence the difference.

Your CNN will have requirements for the type of images it can process. That's completely independent of imshow but if you've not loaded the image properly, then yes you'll run into trouble.

请先登录，再进行评论。

请先登录，再回答此问题。