This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

OPT8241-CDK-EVM: about Lens Calibration for OPT8241

Part Number: OPT8241-CDK-EVM

Hello.

I am studying for calibration now. (finished PCB design)

I've checked that 'Lens calibration' in VOXEL viewer is for distortion correction.

I have one question for calibrationg related with lens.

There was no inform for Relative illumination(RI). I guess there is effect from lens characterstic, RI.

So I am curious if we don't need to take care of RI.

Could you give me any advice? 

  • Hi!

    Lens calibration takes care of the distortion of the image. This means that it takes care of the shape of the objects in the image. However, there is no need to "change" the measurement amplitude of the image. If a part of the image has lower amplitude, it should just be shown as a low amplitude image.
    RI will not affect the depth output so we don't have any calibration for this. If you feel that you would like to correct the amplitude image as well, you would have to look at some external literature.

    Regards,
    Subhash
  • Hi

    Thank you for kind explaining.

    Please help me one more time to make it clear.

    I understand as below.

    I will use lens AOET's 1LS1097F. it's RI is about 49% at corner.

    That means when 100ea photon arrive at center of OPT8241, 50ea photon arrive at corner.

    Even though total photon's Q'ty which arrives at center & corners are different, depth data is calculated with ratio of TAP A & B.

    So lens RI(different Photon's Q'ty at center & corcer) doesn't affect depth data. I fully understand.

    Please check about accuracy.

    I guess as below.

    Because photon Q'ty is less than center's, accuracy of coner is worse than center's.  (using flat wall/chart for measureing)

    In case of normal RGB sensor, circle gain is used for lens shading correction.

    However I think OPT8241 doesn't have any function like this, so cannot handle different accuracy of coner & center.

    Do I understand well?

    (So I think I should decide best ROI for each application,target distance.)

    Thank you.

    I cannot find anyone around me know about TOF. So here is only one good place for me.

  • Hi!

    Your explanation is very good. Indeed, the corner pixels have smaller signal power and more depth noise as compared to the center pixels. It is possible to somehow amplify the signal power at the corners to correct for RI. However, that will not reduce the depth noise at the corners.

    So what you have written makes sense. A good way to correct for this is to use engineered diffusers and modify the light shape of the illumination of the camera. If you project more light near the corner part of the FoV, you will see that the lens receives 100 photons for center pixel and transmits 100 photons to the pixel. But the lens will receive 200 photons for the corner pixel and it will transmit 100 photons to the corner pixel.

    This way, you can "correct" for the Relative Illumination factor of the lens.

    A few vendors for engineered diffusers are RPC Photonics and Luminit.

    Regards,
    Subhash