How can I use the elements of 'PixelValues' of 'regionprops'?

5 次查看(过去 30 天)
Hello,
I have been able to extract the pixel values of the grayscale image using regionprops (I think). I did so by using this code
img = rgb2gray(imread('W1\Writer1_01_02.jpg'));
bw_normal = im2bw(img, graythresh(img));
bw = imcomplement(bw_normal);
[label,n] = bwlabel(bw);
stats = regionprops(label, img, {'Area', 'BoundingBox', 'PixelValues'});
Now I want to extract or see visually the values extracted. I mean with the case of "stats.Area" I can do this to extract its values to another vector
areas = [stats.Area]
But this does not work with PixelValues. If I do the same with PixelValues I am presented with the following error:
>> test = [stats.PixelValues]
Error using horzcat
CAT arguments dimensions are not consistent.
Basically what I am trying to do is to visually plot the pixel values, as in plot a histogram of not the entire image but the PixelValues of the connected components.
This was done using C#. I want to produce something like this. In the image below the pixel values of the connected components is detected and then plotted onto the image. Is this possible? (The red dots are the detected pixel values)
Thank you
  4 个评论
Image Analyst
Image Analyst 2014-2-18
Note that this will give you the same as what I told you
allThePixelValues = img(bw);
Both essentially give you all the gray levels in the binary image that you sent to regionprops, just in a different order.
Faraz
Faraz 2014-2-18
Oh right, I didnt see that comment you added later. Thanks

请先登录,再进行评论。

采纳的回答

Image Analyst
Image Analyst 2014-2-17
You can do that by using imshow on the binary image:
imshow(bw);
If you want to mask it to show the original pixels in grayscale, then you can do that
maskedIMage = img; % Initialize
maskedImage(~bw) = 0; % Non-selected pixels set to zero.
imshow(maskedImage); % Display.
  3 个评论
Image Analyst
Image Analyst 2014-2-17
Then you can just get them from the binary image that you sent into regionprops:
allThePixelValues = img(bw);

请先登录,再进行评论。

更多回答(1 个)

Pedro Marrero
Pedro Marrero 2016-1-10
pixel_value_i = stats(i).PixelValues;

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by