View my account

D435 Gamma Correction

Comments

4 comments

  • MartyX Grover

    Hi Accounts Mmt  I do not have any information on whether the minimum of 100 = no gamma correction at all, unfortunately.  So I would not be confident in saying that it does.

    If you are seeking RGB data that is as raw as possible, it may be worth considering accessing raw bayered RGB images from the camera and using OpenCV to convert them to a usable format.  The discussion in the link below looks at this process and may provide you with sone useful insights.

    https://github.com/IntelRealSense/librealsense/issues/7275 

    0
    Comment actions Permalink
  • Accounts Mmt

    Thank you MartyG. I will look into it.

    0
    Comment actions Permalink
  • Masuda

    I had the same issue with you, and I did a simple calibration by capturing density-varying random dot sheet with various gamma settings with D435i.

    I omit the details, but my conclusion is camera gamma may be 2.5 - k / 300 where k is the gamma setting (100-500). Samples of the gamma values are 2.2(k=100), 2.0(k=150), 1.5(k=300), 1.0(k=450).

    To obtain the calibrated intensity value, apply the inverse gamma like I^(1/camera_gamma) where I is the captured intensity value normalized in 0-1.

    0
    Comment actions Permalink
  • MartyG

    Thanks very much Masuda for sharing your knowledge!

    0
    Comment actions Permalink

Please sign in to leave a comment.