• Difference between rgb initial preview vs. resize?

I noticed in examples there would be a preview size that would be larger (960, 1632) and then it would be put into an ImageManip to get the size down to something smaller (544,320).

Reading through the documentation I cannot figure out how the preview size implementation works. Is it not performing the same process as ImageManip?

  1. I am guessing that starting at (960,1632) enables for a wider field of view vertically and horizontally? Does the preview size of (544,320) crop the image? Then how does setting the resolution come into play with all this?
  2. Does this also apply to the depth image?
  3. If I go preview(960,1632)->manip(544,320) on RGB, and then do setOutputSize(544,320) on StereoDepth, would it have a different field of view and not work on a SpatialImgDetections node? How does StereoDepth resolution come into play?
  • erik replied to this.

    Hi AdamPolak ,
    ImageManip can perform the same as the .preview (if ImageManip settings are configured correctly). Which example are you referring to?

        Hi AdamPolak
        The sensor resolution is what sets the VFOV and HFOV. The preview downsizes the starting resolution if the aspect ratio matches. Otherwise it crops the larger FOV.
        Depth is dependent only on on stereo camera resolutions.

        Thanks,
        Jaka

            jakaskerl

            I am confused how to square that with this information:

            If resolution sets the FOV, then ISP seems to be a part of it? Is there a way to get the "full" view (as shown in the docs below outside of the rectangle)