What causes green-pixel non-uniformity between adjacent pixels on a CMOS sensor?

Asked 5/21/2015

4 views

2 answers

0

On a Bayer CMOS sensor, I’m seeing a consistent imbalance between the two green pixel sites even in flat, evenly lit areas. It looks too localized to be simple CFA variation, but too systematic to be random thermal noise. At a sensor level, what typically causes unequal response between adjacent green pixels? For example, is this mainly due to pixel/column circuitry variation, dark-signal non-uniformity, photo-response non-uniformity, or crosstalk from neighboring red/blue pixels?

Originally by Photography Stack Exchange contributor. Source · Licensed CC BY-SA 4.0

Photography Stack Exchange contributor

11y ago

2 Answers

1

I believe the problem you're asking about is green non-uniformity (GNU) It is typically caused by differences cross talk from adjacent pixels in column direction. The difference in sensitive of the adjacent blue or red pixel will cause differences in the exposure of green pixel in question. This is typically counter acted by maintaining two green gain factors and using algorithms to compensate for the green non-uniformity.

Originally by user38877. Source · Licensed CC BY-SA 4.0

user38877

11y ago

0

AI Answer

Generated from our catalog & community — verify before relying on it.

This effect is commonly called green non-uniformity (GNU). A typical cause is pixel crosstalk: the two green sites in a Bayer pattern sit next to different colored neighbors, so differences in how signal leaks or spreads from adjacent red/blue pixels can make the two greens respond slightly differently even on a uniform area.

In practice, this is treated as a systematic sensor non-uniformity rather than random thermal noise. It can also overlap with general pixel response variation, but the key point is that the mismatch between the two green channels is often spatially structured and repeatable.

A common mitigation is to treat the two green channels separately, applying different green gain factors and correction algorithms during image processing.

UniqueBot

AI

11y ago

Your Answer