D455 - Strange deph measure when use depth to disparity

rs2::disparity_transform depth2disparity(true);
...
measure = frame.get_distance(x, y)

s2::decimation_filter dec_filter(4);
rs2::spatial_filter spat_filter(0.25f, 50, 5, 0);
rs2::temporal_filter temp_filter(0.3f, 20, 3);

// True : Bitmap OK. Depth measures BUGGY.
// False : Bitmap NON OK. Depth measures OK.
rs2::disparity_transform depth2disparity(true);

// Wait frames
rs2::frameset frames = instance_pipeline.wait_for_frames();

rs2::frame image_frame = frames.get_depth_frame()
	.apply_filter(temp_filter)
	.apply_filter(dec_filter)
	.apply_filter(depth2disparity) // true
	.apply_filter(spat_filter)
	.apply_filter(aligner)
	.apply_filter(couleur_depth);

rs2::depth_frame depth_frame = frames.get_depth_frame()
	.apply_filter(temp_filter)
	.apply_filter(dec_filter)
	.apply_filter(depth2disparity) // true
	.apply_filter(spat_filter);

• MartyG

  • December 23, 2020 07:47
  • Edited

  The SDK's post-processing documentation recommends that depth2disparity is applied after the decimation filter and before the spatial filter.  The link below provides the recommended order of applying each filter type.

  https://github.com/IntelRealSense/librealsense/blob/master/doc/post-processing-filters.md#using-filters-in-application-code 

  I would also recommend removing the second round of filter application at the bottom of the script to see if your results improve.

  1

  Comment actions Permalink
• Faviersebastien01

  • December 23, 2020 08:26

  Nice! I'm stupid

  I've re-ordered filters and it's works !

  Thank you very much

  Regards

  0

  Comment actions Permalink

Please sign in to leave a comment.