Mech-Vision 2.1 Release Notes

When an entire point cloud model is used for 3D coarse matching, if the target object was too thin, the downsampling interval could exceed the object’s thickness. This resulted in retaining the point cloud of a single view during matching, causing incorrect matching results.

When the “3D Matching” Step was used as an input within a Procedure and connected to a data flow, the software crashed if the “Input Type” of the “3D Matching” Step was adjusted.

If the “Auto-Set Voting Parameters” option was disabled in the “3D Matching” Step and voting parameters were manually adjusted, executing the single Step would apply the manual settings. However, even after re-enabling “Auto-Set Voting Parameters” and executing the single Step, the manually adjusted parameters were still used for matching.

For the “3D Fine Matching” Step, when the Operation Approach was set to “HighPrecision" and “Speed Up on Large Object Quantities” was unselected, the matching results for symmetric target objects were incorrect.

When a copy of the point cloud model generated by importing an STL file was created in the target object editor, the STL model dimensions became incorrect after returning to the “Configure STL file” step.

After modifying the axis of symmetry for the pick point array in the target object editor, the “Order of symmetry” and “Angle range” displayed in the parameter panel were incorrect.

When a copy of the target object model was created and used in the target object editor, the point cloud of the target object was not removed during project execution, even when the “Remove Target Object Cloud” parameter in the “Output” or “Path Planning” Step was selected.

In the EIH scenario, clicking the Get data button in the pose adjustment tool incorrectly updated the robot flange pose.