(Nice, but without quotes...)

[x,y]=find(A==1);       or

[x,y]=find(A==255);

depending on the pixel-value of white.

you can use the imageJ plugin "point_cloud.jar" which can be downloaed from the link below. Even though it is for CT-scans, it will work with only one-stack image

http://rsbweb.nih.gov/ij/plugins/point-cloud/index.html

if you use MatLab you can use "find" function... type 'help find' for more  details

It seems to be binary images, if it is so, apply connected component algorithm. But first of all, please clarify that what you want to do with these images? Whether you want to extract a portion or doing some processing in internal part of white pixels?

It seems that you don't have metadata for these images (e.g., a header with additional information), so you have to figure out yourself what would be a reasonable resolution. For this, you may first think at the resolution of scanned images (300 dpi, 600 dpi...), and then degrade it, such as the white line on the black canvas is preserved.

Example

Let's suppose that you conclude that you have pixels of 0.5 mm x 0.5 mm, and the canvas (page) is A4 format (210 mm x 297 mm). You define yourself a coordinate system, e.g. "my origin = (0 mm,0 mm) is the Bottom-Left corner of the Bottom-Left pixel", and then you calculate the coordinates (in mm) of all pixels in the current image; each pixel is referred to its bottom left corner {that is, the pixel (x,y) covers an area defined by BL mm=[x+0.5), y ; UR mm=[x+0.5); [y+0.5) }.

Two remarks: (1) once you have informed your images with these relative coordinates (to the Bottom-Left origin) , you can easily both relate your origin, and change the pixel size, to whatever is necessary in your study (other origins and other length units, respectively; in particular, you may link to an animal size, if you need this); (2) you have the coordinates in length units (as described), but you can transform them in pixels [e.g., the pixel "(3,6)" related to the origin pixel "(0,0)" and  you address a matrix, keeping in mind the pixel size if you need to eventually link to the real world].

This is just an example that suggests you a way to approach your problem. It's up to you to finally choose the origin (including a pixel corner out of four, or the pixel center) and units of the coordinate system, as well as the pixel's size, by taking into consideration all information and goals you have. The programming is really easy.

Finally, it's important to be aware  that you always have the information on pixel coordinates in the digital images themselves, either expressed or implied, the latter being your current case.

Like Ruxandra said, you don't have that information. It is not present in a bmp, tiff or jpg.

You can use pixel-indices as coordinates, but you could also scale it if you want. If so, rotate the image so that the cow stands on four legs, and measure the distance from head to tail in number of pixels. If you know how big the cow is, you can derive a pixel-size. Choose an arbitrary origin and arbitrary direction of +X and +Y, and you are done!

i want the coordinates of X and Y were the value is one (1).

for example as shown below, if this is the matrix of the image, i want the coordinates of x and y were value is one (1).

0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 1 1 1 1 1 0 0

0 0 1 1 1 1 1 1 1 1 0

0 0 0 1 1 1 1 1 1 1 1

0 0 0 0 1 1 1 1 1 1 0

0 0 0 0 0 0 0 1 0 0 0

0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0

That's exactly what the plugin 'point_cloud.jar' do. It writes a txt file with a header like "Cartesian Coordinates: X Y Z"

As you only have one slice z=1 and x-y will be the coordinates of pixels with the desired values (255, for example)

http://rsbweb.nih.gov/ij/plugins/point-cloud/index.html

Umar, do you want to know the matrix coordinates of all the pixels that has value=1?

@Parul Nilesh Shah yes i want the coordinate x and y at which the pixel value is 1

By convention pixel-indices start at the upper-left corner, and start with zero. For the first 1 in your example, the pixel-index would be 4 + 2 * DimX , or -expressed as (x,y)-, (4, 2) . All in arbitrary units, because the picture is not calibrated.

Here's the C- way of doing it (simple and obvious, my preference ) do not bother about the complexity of computation and other overheads ...

 cnt=1;

     for (i=0;i

if you use MATLAB you can implement this code...

a=rgb2gray(imread('PostFileLoader.tif'));

[x,y,z]=size(a);

count=0;

for i=1:x

  for j=1:y

    if a(i,j)~=0

      count=count+1;

      new_x(count)=j;

      new_y(count)=i;

    end

  end

end

You can get the coordinates values from new_x & new_y

if you are working in matlab, simply use the "Pixel List" function to extract the image coordinates.

I think what Sumi has suggested should work for you.  Basically identify the size of the image using 'size' command and then scan the matrix one pixel at a time by incrementing row and/or column index. Wherever you find pixel value = 1, the respective index is what you want.

Good luck

if A contain your binary image simply use this command in matlab

[x,y]=find(A==1);

(Nice, but without quotes...)

[x,y]=find(A==1);       or

[x,y]=find(A==255);

depending on the pixel-value of white.

Thank you Lambert for correction

% bwImage is the binary image

%Extract linear indexes of foreground pixels and then convert it to subscripts

[x,y] = ind2sub(size(bwImage),find(bwImage));

thank you all

Please, check this link

http://www.mathworks.com/help/images/getting-information-about-the-pixels-in-an-image.html

If you are using ImageJ, and your image is binary, just go to Analyze -> Tools -> Save XY Coordinates. it will output a txt file with the horizontal and vertical coordinates of each pixel with value 0