"cropping" is no usual term, thus let me try to answer in my terms:

If cropping means to 'cut off' some areas-of-no-interest (along the edges) - without resampling etc. - this will not change the pixel values.

Another way to achieve the same would be to copy the area-of-interest to a new image of smaller size - pixel by pixel.

One restriction:

You are talking about color images. The above is valid for images that are already represented by e.g. RGB tripel pixels. Images still in the sensor format (e.g. Bayer-patterned) will show some aberrations along the edges if de-mosaicing is performed post 'cropping'.

Hello Hari,

If you crop the unwanted region from a image. There will be no changes in in the pixel value however you can noticed the the difference in intensity of the color visually.  But if you compare  pixel of the interested area before cropping and after cropping, the pixel value will be same.

Regards,

Hi Hari,

As Ranjit said, the pixel values would not be changed after cropping operation. cropping operation will not do any re-sampling operations on the image. It will not expand or reduce the size of the image. Thus, the area of interest is being taken out by using cropping operation.

Thanks & Regards,

Sumaiya. M.N.

 Hello Dear

Changing the image size  means scaling the whole image (or just a selected region) up or down. Resizing is actually a kind of distortion because the image content is being stretched or squashed.

Changing the canvas size  just involves adding or taking away pixels around the edges of the image. It's like adding to the neutral border around a mounted photo, or taking a pair of scissors and cropping the photo to a smaller size. In either case, the remaining image pixels are undisturbed so there's no distortion.

hi friend, 

apart from the above explanation, the reduced image cann't be used to compared with ground truth for exact research in my perceptions 

Thank you all for your explanations.

Let me explain in detail my problem. I have a CCD captured image in which only center of the captured area contains the main image. All the remaining area is dark except the center region. Now i want to extract only that part where the main image is located. so i cropped the image and using the pixel data for further analysis.

My doubt is whether the cropping operation is reliable or not? Means the cropping will alter the pixel value or it remain unchanged? 

Extracting (copying) the area-of-interest doesn't change the pixel value.

Yes offcourse you can go for further analysis with the cropped image. Its 100 % reliable.

If the image is not compressed (raster) there should be no differences in the selected pixels before or after the crop.

If you bring a compressed image (i.e. a .jpeg) you may find little differences since your crop may affect the selection of the 8x8 blocks of pixels used for the DCT compression.

Cropping the image (without resampling or other processing after the fact) will certainly retain the number of pixels used to depict the part of the image you are interested in retaining, yes. Obviously, the overall pixel count in the frame will be reduced, from what it was in the original frame. You have discarded a bunch of pixels that were of no interest.

So, you will not lose information in the area of interest. Is this what you were asking?

If the original frame was, say, 1920 X 1080 pixels, but you are only interested in the middle part of the picture, you can certainly crop the image into a smaller, say, 720 X 480 pixel image, and discard the rest.

I chose those pixels counts to show what happens to resolution. The initial image was by convention "high definition" (HD). By cropping away the borders, and retaining only the middle pixels, you are now left with a "standard definition" (SD) image frame.

However, the resolution of that middle part of the image has not changed.

Let's assume you have a high definition display. If you fill your display with the original 1920 X 1080 frame, you will see a sharp image, with lots of content in it. (Well, in your specific case, mostly black borders. So assume instead a large image with very much content.)

If you then fill that same display with the cropped image, expanding the cropped 720 X 480 image to fill the entire display, the result will be a softer, SD image rendition. The display electronics will upconvert the SD image to fill up the HD display area, however this process will not create detail that did not exist in the central part of the original image. So, the result will appear softer overall, even though you haven't lost info from that central area of the original.

This effect is analogous to what happens with conventional silver halide photography. The grains used to depict a small part of the negative remain the same, if you crop and then enlarge the picture. When you look at the cropped and then enlarged picture, the overall effect will be "more grainy" than the uncropped image would have been, on the same size final print.

Thank you very much sir for your well elaborated explanation. @Albert Manfredi.