Sunday, February 01, 2015

Sony Proposes Photon Counting Pixel

Sony patent application US20150021461 "Image sensor and electronic device" by Toshiyuki Nishihara and Hirofumi Sumi proposes a buried channel TG and binary detector to reduce influence of TG surface traps in very low light image sensors:

"since a pixel signal is extremely weak when weak light is detected, it is desirable to reflect electrons generated from photoelectric conversion on intensity of the pixel signal while losing as few electrons as possible. In general, however, when electrons generated in a photodiode are transferred to a floating diffusion, carriers (electrons) are doped at an interface level generated due to a defect present on a gate oxide film interface (interface defect) of a transfer transistor."

"there is provided an image sensor including pixels each configured to include a transfer transistor configured as an embedded channel type MOS transistor and to output a pixel signal based on a charge transferred to a floating diffusion from a photodiode by the transfer transistor in an on state, and a determination unit configured to convert the output pixel signal to a digital value, then compare the converted digital value to a threshold value, and thereby make a binary determination on presence or absence of incidence of a photon on the pixel that has generated the pixel signal. Accordingly, the image sensor that makes a binary determination on presence or absence of incidence of a photon on a pixel exhibits the effect of reducing influence of an interface state on the transfer transistor."

10 comments:

  1. Well, well, well...suddenly I have company on this explorative journey. Welcome Sony to the land of QIS.

    ReplyDelete
  2. I think that this approach is mainly for long exposure time rather than exploring QIS concept.

    ReplyDelete
  3. This is great news for imaging. It means we'll see this technology come to market in our lifetimes rather than staying stuck in theoretical research papers. It also means the intellectual property will have a chance to blossom in the industry rather than be subverted by an aggressive patent strategy.

    ReplyDelete
    Replies
    1. How does Sony IP give the technology a chance to blossom in the industry?

      Delete
    2. I think its more because Sony is an actual manufacturer and not a research institution, or worse, a patent troll. Even with promising research finding someone to manufacture and market a product needs a convincing case which will be less of a hurdle at a company such as Sony who has been known to test the waters with unproven technology.

      Delete
  4. Very interesting approach ! The electronic noise accumulated during long exposure time is a killer for ultra low light photography. With photon counting method, you will only increase the counter when a photon is detected, so the electronic noise under the threshold will not be accumulated. The price to pay is a lower DQE but the benefice is great at ultra low light conditions. Taking image intensified sensor as example, the QE of the photocathode is very low compared to silicon device, but the high gain removes virtually all the electronic noise and gives the best low light performance in many cases.

    -yang ni

    ReplyDelete
    Replies
    1. I am not sure what you mean by electronics noise accumulated during long exposure times? Do you mean dark signal? Unfortunately we cannot discriminate between dark electrons and photo electrons in such a system so no help there. But gain is definitely helpful, which is an advantage of EMCCDs and SPADs.

      Delete
    2. Dark current is part of the electronic noise. If you do multi-frame accumulation, then all the thermal noise, switching noise, etc in the readout chain will be accumulated. When you make an early decision on the presence of single photon electron, then you go to symbolic domain.

      -yang ni

      Delete
    3. If we remove dark signal and glow signal from consideration, and consider just a regular PPD but with high conversion gain, I don't think a long integration time matters much in terms of read noise. If you mean digital integration, where read noise accumulates over many reads, that is a different story, and there low read noise and a quantizer helps, but analysis gets more complicated. That is why I moved to bit error rate (BER) for the single bit QIS analysis.

      Delete

All comments are moderated to avoid spam.