General Settings
After bin recognition is completed, you can configure auxiliary functions other than visual recognition in this workflow. Currently, only output port content configuration is supported.
Based on actual needs, you can choose whether to output target object data inside the bin. After selecting the corresponding options, the "3D Bin Recognition" Step adds corresponding output ports in real time.
-
Select output port data.
-
Depth image and color image: Output the depth image and color image inside the bin.
-
Surface point cloud and edge point cloud: Output the surface point cloud and edge point cloud inside the bin.
-
-
Select data calculation method.
-
Calculate mask difference: Output target object data inside the bin by calculating mask difference between bin and target objects. If the top surface point cloud of the bin is complete, this method can be used to obtain target object data inside the bin.
-
Extract data by bin dimensions: Extract target object data inside the bin based on bin dimensions. If bin dimensions are accurate, this method can be used to obtain target object data inside the bin.
-
Target object and bin segmentation (recommended): Based on the target object and bin segmentation model package, segment the input depth image and color image into target objects and bin, and obtain target object masks and bin mask. If there are many interference points at the bin bottom, this method can remove interference points.
If this method is selected to calculate target object data inside the bin, make sure the "Target object and bin segmentation" feature is enabled and related parameters are set in the "Point cloud preprocessing" workflow.
-
Now the configuration in the 3D Bin Recognition Tool is complete. Click Save to save the configuration.