Dear Mala,

You can test the following source code:

from PIL import Image

x=5 # x coordinate you want

y=5 # y coordinate you want

im = Image.open("name.jpg")

pix = im.load()

print pix[x,y]

Not Python 3.5, but here you have a well documented tutorial about image processing using Python 2.7.X and OpenCV: 

http://opencv-python-tutroals.readthedocs.io/en/latest/py_tutorials/py_gui/py_image_display/py_image_display.html#display-image

Dear Evagoras,

Thanks a lot fr the code. I tried this but the printed x, y values are not giving the intensity. Reading the header is very important. Header contains the pixel no.s in x and y direction. Using these values all the pixels can be read and processed. I have not seen any code for reading the header.

Thanks again

--Mala Mitra

Dear Marco,

Thanks for the link. The link gives lots of info. However, what I really need is to read the intensities of a pixel and write / save the modified / processed value in a new file. Image header processing should be done.

Thanks again..

Dear Evagoras,

Many many thanks for your code. Only a small typo is to be rectified. It is: print(pix[x,y])

Your code:

********************************

from PIL import Image

x=5 # x coordinate you want

y=5 # y coordinate you want

im = Image.open("name.jpg")

pix = im.load()

print pix[x,y]

*******************************

Earlier I was making a mistake in the PIL import command. Any idea, how to get the size of the image? I mean max x and max y. This needs to be obtained from the header. I want to select your answer as the correct answer but that button is now missing.

Thanks again

--Mala Mitra

In both PIL and opencv (imread) you (can) get a numpy array as a result.

There you just ask its "shape" =(height,width,channels).

Hi Geert,

Thanks a lot for your response. One small example code will be very helpful.

Dear Mala,

Please find in attached file (Image.py) a short sample of python code to read an RGB jpeg image, find the shape and store red, green and blue channels into numpy arrays.

I use imageio library instead PIL, Pillow or open_cv but the "imread" function from imageio is easily replaced by another.

The image use with this example is in attached file (RGB_Image.jpg).

There is a link to the Matplotlib image tutorial : https://matplotlib.org/users/image_tutorial.html#importing-image-data-into-numpy-arrays

I hope that this example will help you.

https://matplotlib.org/users/image_tutorial.html#importing-image-data-into-numpy-arrays

 Dear Francois,

Thanks a lot for the code and the tutorial. I executed the code in Spyder 3.5. There is an error: "ImportError: No module named 'imageio' ".

I did a little bit of modification as given in the tutorial and it worked. However, I do not want to plot. I want to get the size of the image from the header. The Python code should read the header and return the # pixels in x and y directions. Then I can read the RGB values of all the pixels, process and store it.

Best Regards

--Mala Mitra

Dear Mitra,

I do not understand if you want to get the header information or if you want to know the size of the image in pixels.

To know the size of the image in pixels, you get it with the command (see my proposed code):

x, y, channel = (RGB_Image.shape)          # X size, Y size, number of channel (3 for RGB image).

 To get the information from the header, it depends on the file format.

Best Regards,

Francois

from PIL import Image

this line don't get execute the file. Error showing everytime. Define issue of import command.