How to extract the value pixel values from an image or masked image?

423 views (last 30 days)
I need the values of the pixels from an image. I need values for each pixel separately.
  3 Comments
Image Analyst
Image Analyst on 11 Jan 2021
This is so vague, just like the original question. If you have read in your image with imread() or gotten it some other way, like from a video camera, then you have your pixels already and you just need to reference them like the Answers below. Please explain in detail why the answers below do not give you any pixel values.
DGM
DGM on 12 Feb 2023
Oh that's easy. Say you have an image called myimage. Then to detect pixels:
haspixels = ~isempty(myimage);
You're welcome.

Sign in to comment.

Accepted Answer

Image Analyst
Image Analyst on 21 Aug 2014
Not sure what you mean. The image itself is a collection of pixel values. To get the pixel value at one particular (row, column) location, you can just specify the index:
grayLevel = grayImage(row, column);
or you can use impixel():
rgbColor = impixel(rgbImage, column, row);
  7 Comments
Claudio Ignacio Fernandez
Claudio Ignacio Fernandez on 16 Jul 2020
Hello @Image Analyst
Today I used ginput over my image to get the x y coordinates. However, instead getting for example an x y values [40 80] I'm getting [40.12 80.07]. Why this is happening? any lead?
Anyway when using the weird X Y coordinates with the command impixel I get pixel values, however How can I know if I'm getting the value of the pixel I really want??
Image Analyst
Image Analyst on 16 Jul 2020
That's the way ginput works - it gives you floating point values. You need to round to get array indexes of row and column
[x, y] = ginput(1);
row = round(y);
column = round(x);
Remember that row is y and column is x so don't make the mistake of saying yourImage(x, y) to reference pixels -- it's yourImage(y, x) which is yourImage(row, column).
pixelValue = yourImage(row, column); % If it's a gray scale image.
pixelValue = yourImage(row, column, :); % If it's an RGB Color image. pixelValue will be a 1x3 vector of [r,g,b] values.

Sign in to comment.

More Answers (6)

Ben11
Ben11 on 21 Aug 2014
You can get the histogram of pixel values using imhist.
For example:
1) Grayscale image
Image = imread('coins.png');
[count,x] = imhist(Image);
2) RGB image
Image = imread('peppers.png');
[count,x] = imhist(Image(:,:,1)); % select one of 3 channels
or use rgb2gray:
Image = imread('peppers.png');
[count,x] = imhist(rgb2gray(Image));
  6 Comments

Sign in to comment.


yonatan gerufi
yonatan gerufi on 21 Aug 2014
Hi Dhanya,
you can access to a specific pixel by typing : figure_name(x_pos,y_pos) .
In the MATLAB workspace, most images are represented as two-dimensional arrays (matrices), in which each element of the matrix corresponds to a single pixel in the displayed image. (from Matlab documentation )
This matrix can be represented in several types as double, uint8, uint16. It can also be RGB, intensity, or indexed types.
I highly recommend reading to understand the differences.
  8 Comments
Guillaume
Guillaume on 24 Oct 2019
Edited: Guillaume on 24 Oct 2019
My comment was addressed to Gustavo whose code is just a more convoluted
numPixelsInImage = numel(riceImage);
His length(array(1:end)) (which is the same as length(array(:)) by the way) reshapes the image into a vector just to know how many pixels there are. Indeed it (and numel) only works with grayscale images.
And, yes reshape does not reorder the pixels. However, it does use some cpu cycles for a completely unnecessary operation.
Image Analyst
Image Analyst on 24 Oct 2019
I was told by a Mathworker that (:) does not reshape the array into a column vector. So passing it into size() or length() would not reshape it. He said that if you assign that to a NEW variable, that new variable will be of a columnar shape, but that is a new variable and there is no temporary variable that is created with a column shape nor is the original array reshaped into a column vector. Only a new variable would have that shape. Fine point though.

Sign in to comment.


Youssef  Khmou
Youssef Khmou on 21 Aug 2014
Accessing a pixel is similar to retrieving element from matrix, here are two examples :
for gray scale image :
X=imread('circuit.tif');
X(10,60)
for multi channel image :
Y=imread('autumn.tif');
Y(10,60,1) %R
Y(10,60,2) %G

snehal jaipurkar
snehal jaipurkar on 23 Nov 2016
after finding the count of each pixel value seperately, i want to add the counts of a range of pixel values and get the total, how to write its code in matlab??
  2 Comments
Image Analyst
Image Analyst on 10 Nov 2018
Use sum:
sumOfCounts = sum(counts(index1:index2));
where index1 and index2 define the "range of pixel values" that you want to sum over.

Sign in to comment.


Umar Awan
Umar Awan on 25 Feb 2019
how can i convert an 28*28 pixel image into 1*784 ? means in 1 row
  1 Comment
Image Analyst
Image Analyst on 25 Feb 2019
If grayImage is your 2-D image, do
rowImage = reshape(grayImage, 1, []);
to reshape the 2-D gray scale image into a 1-D row vector.

Sign in to comment.


Asad Alam
Asad Alam on 25 Feb 2021
How can i compare pixel value of an image
pixelvalue<300
And i want all the pixels whose values are above 300. can anyone help
  2 Comments
Gustavo Liñan
Gustavo Liñan on 25 Feb 2021
% Since you're specifying a single scalar threshold, I assume your image is grayscale
% If you want a binary of the same size just use, assuming your image is already in memory
% in variable YourImage;
YourThreshold=300;
RESULT=YourImage<YourThreshold;
%If your image is in a File;
YourImage=imread(fullfile(YourImagePath,YourImageFileName));
RESULT=YourImage<YourThreshold;
%Now if you want this result as a vector of Ncols x Nrows elements, just do
RESULT_VECTOR=RESULT(1:end);
%The indexes for these pixels are simply
MY_PIX_IDX=find(RESULT_VECTOR)
%Now if you want to extract the pixel values satisfying the condition
YOUR_PIXELS_BELOW_THRES=YourImage(MY_PIX_IDX);
The question is indeed a bit vague to give a completely operative solution, hope these hints help.
Image Analyst
Image Analyst on 25 Feb 2021
Edited: Image Analyst on 25 Feb 2021
This is not an answer to @Dhanya and should have been it's own question. Anyway...
To get a binary image "map" of where those pixels are, you can do
binaryImage = yourImage < 300;
Now these might be some irregularly-shaped "blobs" as we call them in image processing. To extract those pixels in the blobs into one giant list (vector), you can do
listOfPixelValues = yourImage(binaryImage);
If you want the values of each blob separate from all other blobs, you need to call regionprops
props = regionprops(binaryImage, 'PixelValues');
props is a structure array. You can also ask for the data to come back in a table instead of a structure array if you want.

Sign in to comment.

Community Treasure Hunt

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

Start Hunting!