Should I use depth_intrinsics or color_intrinsics for rs.rs2_deproject_pixel_to_point if If depth is already aligned to color?

[monacv]

If I already have aligned depth frames to color (RGB) frames, for using the rs.rs2_deproject_pixel_to_point method, should I use depth_intrinsics or color_intrinsics and why?

[MartyG-RealSense]

Hi @monacv Can you confirm please if you have performed depth to color alignment on 2D depth and color frames, and that you then want to use rs2_deproject_pixel_to_point to convert an aligned 2D image to 3D points, please?

My understanding is that if you are performing alignment then the intrinsics that you use are those of whatever you are aligning to. So if you are aligning depth to color then color_intrinsics is used. This reflects how, when aligning depth to color, the depth FOV is aligned to the color FOV size.

The discussion in the link below is a good reference that covers the concepts of depth to color alignment, rs2_deproject_pixel_to_point and color_intrinsics.

#1413

[MartyG-RealSense]

Hi @monacv Do you require further assistance with this case, please? Thanks!

[MartyG-RealSense]

Case closed due to no further comments received.