Options — 3D
The 3D tab configures volume measurement: where the conveyor's "zero plane" sits, how the point cloud is built, and — if a secondary depth camera is configured — how it's aligned to the top camera. It only applies when Enable 3D Vision is turned on for the active Vision Configuration.
A live point-cloud viewport on the left shows the effect of every setting on this tab immediately.
Zero Plane
Defines the conveyor's zero plane: its height (Z offset) and tilt. This is saved as a normal + offset and used later to voxelize segmented objects and calculate their volume — it's not just a visual overlay, it directly affects measured volume.
| Field | Range | Description |
|---|---|---|
| Z Offset | 200–1000 | Height of the zero plane. |
| Rotation X / Rotation Y / Rotation Z (deg) | -5–5 | Tilt of the zero plane on each axis. |
| Voxel size (mm) | 1–20 | Cell size of the volume grid on the zero plane. Smaller follows the shape more closely; larger smooths over holes in sparse clouds but overestimates at the product's edges. |
Depth/Top Camera Alignment
Only shown once a secondary depth camera is enabled. The point cloud is colored by sampling the top camera's image through this mapping: depth_x = top_x × scale X + offset X (same formula for Y).
To align: put an object on the belt, switch the viewport to RGB, and adjust the sliders until its colors sit on its shape. The magenta rectangle in the viewport shows where the top image lands at belt height; grey is what the depth camera covers; dark grey points fall outside the top image altogether.
| Field | Range | Description |
|---|---|---|
| Alignment scale X / Alignment scale Y | -2–2 | A negative scale mirrors that axis between the two cameras. |
| Alignment offset X / Alignment offset Y | -1000–1000 | In depth-camera pixels. |
Select Auto align to fit all four values automatically from a color capture off the depth camera — it also reports when the mapping needs a rotation these four values can't express. Until alignment is set, a warning reminds you that no usable depth will be measured (the default scale 1 / offset 0 assumes the two cameras see exactly the same image, which is almost never true).
Depth Post-Processing
Unlike the zero plane above, these change the measured point cloud itself — volume results move with them, not just the preview.
| Toggle | Description |
|---|---|
| Hole Filling | Fills pixels the sensor could not measure from their nearest measured neighbor. Produces a denser cloud, but the filled pixels are estimated rather than measured. |
| High Accuracy Profile | Discards low-confidence pixels instead of reporting them. Less noise on what survives, but a markedly sparser cloud. |