find the coordinates of an image

how to get the xy coordinates of palm from the image

[John]

If you expect to get an answer you will have to be more specific. The picture on the right looks like a MATLAB binary image and all of the "true" pixels represent the palm. Do you want to find the "centroid?" The bottom-left-most pixel of the palm? The bottom-right-most pixel of the palm? The top-left-most? See what I mean?

A good first thing to do would be to specify the problem exactly.

[RAMESH MUNIRATHINAM]

thank you, consider only the right image, i want to get the x-y coordinates of entire structure of palm alone, i know that if we have the coordinates we can get the image, but i have image and i want to know the coordinates is it possible?

[Mohammed Qusay]

@John ... i want to the The bottom-left-most pixel of the palm, The bottom-right-most pixel of the palm,The top-left-most and the top-right-most of the binary image how to get it ?

Answer 1

[Geoff Hayes]

Ramesh - if the right-hand image is a binary image of 0's and 1's, then can't you get the rows and columns (y and x coordinates respectively) using the following

 [rows,cols] = find(binaryImg==1);

where binaryImg is your binary image and 1 corresponds to the white pixels?

[RAMESH MUNIRATHINAM]

THANK YOU GEOFF, i need to model the palm, so with the binary values i can reconstruct the image using matlab, but is it possible to model using binary 1's and 0's.

if i get the coordinate values i will export it to some files and use it for modelling.

[Geoff Hayes]

The rows and cols variables will give you the coordinates of the palm (and fingers). You can then use those coordinates to extract the palm data from the original image.

[RAMESH MUNIRATHINAM]

thank you Geoff, will try and let u know.

[Image Analyst]

If, by chance, you want just the outline of the palm, you can use

boundaries = bwboundaries(binaryImage);

Or (not as efficiently)

[rows, columns] = find(bwperim(binaryImage));

Geoff's code can skip the unnecessary step of comparing every pixel in the entire image to 1 by simply just doing:

[rows,cols] = find(binaryImg);

[RAMESH MUNIRATHINAM]

@image analyst

thank you will inform you after trying it