Minimum size for depth calculation

• MartyX Grover

  • April 13, 2022 09:03
  • Edited

  Hi Redyoung  In the link below, a RealSense team member advises that objects be at least 9 pixels in size in order to be trackable.

  https://github.com/IntelRealSense/librealsense/issues/4175

   

  The D455 model has a minimum depth sensing range of 0.4 m / 40 cm.  When objects / surfaces are closer to the camera than that distance then the depth detail will progressively break up as the object / surface moves increasingly nearer the camera.

   

  You can enable the camera to get nearer to the object by reducing its minimum distance with the Disparity Shift option.  Increasing the Disparity Shift value from its default of '0' to a higher value will reduce minimum distance, though also reduce the maximum observable depth sensing range at the same time (though this is not a problem if performing close-range depth sensing).   A good balance between close range sensing and retaining background detail is provided by a value of '50' for the Disparity Shift.

   

  For thin objects, you may be able to achieve an improved depth image by reducing the D455's default depth unit scale from its default value of '0.001' to the smaller '0.0001' value, as demonstrated in the case at the link below when depth-sensing a toothpick.

  https://github.com/IntelRealSense/librealsense/issues/8228

   

   

  If changing camera models is an option for you, Intel recently introduced the new D405 model for very close range accurate depth sensing, with an ideal depth range of 7 cm to 50 cm.

  https://store.intelrealsense.com/buy-intel-realsense-depth-camera-d405.html

  0

  Comment actions Permalink

Please sign in to leave a comment.