Signal Processing Stack Exchange is a question and answer site for practitioners of the art and science of signal, image and video processing. It only takes a minute to sign up.
Sign up to join this community
Is there a difference in SNR (or any other measure of image quality) between using 2x2 pixel binning in the camera and shrinking the full resolution image in software (cv::INTER_AREA) to the same size (w/2, h/2) ?
I imagine that SNR is higher with binning if pixel binning is done "in hardware" but I'm not sure how binning is actually implemented or if it does make a difference, to what extent the difference is measurable/negligible.
I assume you mean Binnning at the Pixel Level before Demosaicing (Bayer Pattern) and resizing image after Demosaicing.
The main difference has to do with the properties of the Noise.
Demosaicing creates spatially correlated noise which means "Averaging" becomes less effective in reducing it. At RAW level noise is much whiter hence averaging is more effective.
Yet the real advantage of Binning is creating "Super Pixel" which contains all colors in a single element. Hence less demosaicing artifacts.
If you have access to the raw pixel data, you can implement binning in software too. They should be identical upto the step-size of the ADC, by which the quantized analog sum of pixels may differ from sum of quantized pixels.
Hardware binning has the advantage that it has to transfer less number of pixels, hence can operate at high fps compared to raw data transfer.
With hardware binning I believe you get 4x the signal and 1x the readout noise, adc noise etc.
With sw scaling you are working on pixels that have 1x the signal and 1x the noise.
