The Python script is available in '$VISP_WS/visp/tutorial/tracking/model-based/generic-rgbd-blender/' folder.

ground truth in 'ground-truth/Camera_L_%04d.txt' files.

rendered depth images in 'depth/Images%04d_R.exr' files.

rendered color images in 'color/%04d_L.jpg' files.

homogeneous transformation between depth and color camera in 'depth_M_color.txt'.depth camera intrinsics in 'Camera_R.xml'.color camera intrinsics in 'Camera_L.xml'.animate the RGB-D camera and retrieve in '/tmp/teabox' folder.Here we want to run get_camera_pose_teabox.py Python script inside Blender to: Note The project used to create this tutorial is available in $VISP_WS/visp/tutorial/tracking/model-based/generic-rgbd-blender/data/teabox/blender/teabox.blend. We strongly recommend that you save your project. This completes the configuration of the scene in Blender. In 'Object Mode' (1) select the 'Cube' the right pannel (2).Open Blender and do the following to transform the cube in a box with the expected dimensions: Here we will show how to create with Blender a box with the following dimensions: In this section, we will create the corresponding tea box, RGB-D camera and define an initial and final RGB-D camera pose used to animate the scene. The relations that links the pixel size with the focal lenght, the camera sensor size and the image resolution are the following: Note In ViSP (see vpCameraParameters class) and in the computer vision community, camera intrinsic parameters are the following:

Both cameras are grabbing 640 x 480 images. The distance between left and right cameras is 10 cm. We will also consider a RGB-D camera where the left camera is a classical color camera and the right camera a depth camera. Width (from point 0 to point 7): 0.068 meters.Length (from point 0 to point 3): 0.165 meters.Height (from point 0 to point 1): 0.08 meters.Analysing teabox.cao we can see that the box dimensions are the following (see teabox.cao example):

For ios instal Blender 3D 3.6.4