when compressing this image

(Fig2.20 from the book Digital Image Processing by Gonzalez and Woods)

(It corresponds to Fig2.20 from the book Digital Image Processing by Gonzalez and Woods and the original .tif document can be downloaded through this link

By reducing the intensity resolution with Matlab I obtain a similar image that exaggerates the following fluctuations brightness fluctuations in the background. enter image description here

Just out of curiosity, does anybody know what the origin for those is? I am puzzled as to whether they originate

at a physical level, like small stains on the lens that took the photograph, (although they suspiciously stop near the watch)

at the transition of the signal into digital form when the image is captured by the pixels of the camera for example

at some step of compressing/ converting the image from one format to the other to obtain the .tif image. Thanks for your help!

  • $\begingroup$ first observation is: your uncompressed image was badly resized; you can see aliasing artifacts all over the place. $\endgroup$ Jan 29 '18 at 13:03

I've taken the original tiff and just set a low threshold for 1bit quantization (conversion to black and white):

quantized image

so we can clearly make out that these artifacts look very circular, and placed in specific patterns around the image borders.

I think these are artifacts could stem from a resizing of the image with imperfect anti-aliasing, though I'm certainly confused by the shape and the lack of "peaks" along the x and y axes. That would actually suggest lens artifacts, as these tend to be circularly distributed. But I think we can rule this out, since the "dot array" at the edges of the picture clearly indicate this happened after cropping the picture digitally.

Another interesting find: quantizing at the other end yields this:

quantized 2

Looks like "cathode ray burn in" used to look – I would, however, not know how this could happen with a digital camera or a scanner. Maybe it's actually a feature of the face of that clock?

  • $\begingroup$ Wow thanks for the help but what does it say on the upper part of the watch? ... with concern... that fascinates Thomas... a new lease... $\endgroup$
    – good_one
    Feb 6 '18 at 11:56

Maybe at some point the image was reconstructed as stippling circular brush strokes by a greedy algorithm. Greedy, because in Marcus's high-contrast version of the image it looks like some darker circles were painted over with lighter-colored circles, and then again by darker circles.

Second thought, I think someone simply painted over the background with different size transparent circular very soft-edged brushes to make it of flat color. This becomes more apparent by adjusting the contrast like below. I recall that at least in older graphics software, a large brush would not travel smoothly but jump in steps.

Contrast adjusted


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.