1
\$\begingroup\$

Question in brief: at a sensor level what are the typical causes of an unequal response between adjacent green pixels?

Background: I'm experimenting with a sensor, which is typically deployed in the camera unit of mobile phones,tablets etc, and I can see there is an imbalance between green pixels even on flat areas. It seems too localised to be caused by CFA variances, so I'm trying to understand what is causing this at a fundamental sensor level. Equally the effect seems too global to be caused by thermal noise components. My first thought was this could be caused by an imbalance in the column and row transistors that are present in a typical CMOS pinned photodiode sensor - I guess this would be a dark-signal non uniformity

Matt Grum suggests it could be PRNU - maybe you could provide more info why you think this?

Otherwise any comments, suggestions or links to further reading material would be greatly appreciated.

Thanks

Improve this question
\$\endgroup\$
5
  • 1
    \$\begingroup\$ Sounds like an instance of PRNU (photo response non-uniformity). \$\endgroup\$
    – Matt Grum
    May 22, 2015 at 11:30
  • \$\begingroup\$ What do you mean by a "Mobile sensor"? Do you mean one in a mobile phone? \$\endgroup\$
    – Digital Lightcraft
    May 22, 2015 at 15:29
  • 1
    \$\begingroup\$ 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 expiate 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. \$\endgroup\$
    – agf1997
    May 27, 2015 at 16:32
  • \$\begingroup\$ @agf1997 that seems very plausible. If you can add this as an answer, I'm happy to mark it as such. I would like to understand what causes the differences in sensitivities? Thanks \$\endgroup\$
    – trican
    May 27, 2015 at 22:29
  • \$\begingroup\$ @trican I would but the question has been placed on hold. \$\endgroup\$
    – agf1997
    May 28, 2015 at 2:34

1 Answer 1

Reset to default
1
\$\begingroup\$

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.

Improve this answer
\$\endgroup\$
1
  • \$\begingroup\$ Great - one small request before I mark this as the answer; would you mind just clarifying the point of different sensitivities - is this related to the different absorption depth of red versus blue or is it something to do with additional electrical circuits (perhaps as part of the readout circuitry). \$\endgroup\$
    – trican
    May 28, 2015 at 19:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

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