Questions about the main camera and ML camera API

I was testing the native camera examples and I have a few questions regarding the main/cv camera:

  1. Undistorted Frame from the Camera:
    Currently, I'm receiving a distorted frame along with the distortion parameters in the intrinsic parameters. Is there a way to obtain an undistorted (rectilinear) frame directly from the camera, or is it necessary for me to handle the undistortion myself?

  2. Getting the Raw Camera Output:
    In the attached image of an ArUco marker which I captured by saving the frame buffer directly to a file (using YUV_420 profile), there is a white halo around the black edges (unsharp mask effect).
    Is there a way to get the raw frame without the filter that applies this effect?

I'm using ML2 OS and SDK version 1.7.0

Hi @yonatan Welcome to the Magic Leap Developer Forum !

  1. We do not provide the undistorted camera frame, so you will need to undistort the image manually.
  2. Do you mind sharing more information on how the frame was captured and at what resolution? The camera does not apply any filtering. The artifacts are a limitation of the camera resolution and the sensor itself. That said, some users reported that their images were sharper when they used the Capture image functions rather than using the Video Frames directly.