Select Configuration Workflow for Target Objects

You are currently viewing the documentation for the latest version (2.1.2). To access a different version, click the "Switch version" button located in the upper-right corner of the page.

■ If you are not sure which version of the product you are currently using, please feel free to contact Mech-Mind Technical Support.

You can refer to the table below to learn about various workflows for target object configuration and choose the appropriate one based on the actual application scenario.

Basic operation procedure	Configuration workflow displayed in the software	Application scenario	Example
Import STL file to generate point cloud model and configure pick points manually	Import STL model	The consistency between the point cloud acquired by the camera and the STL model file is high. The diversity of incoming target object orientations, along with the changes in the target object pose, will lead to significant changes in the features on the target object’s point cloud. The surfaces of the target objects are reflective, and the missing areas in the point clouds vary between different objects. Scenarios for picking neatly arranged objects. Scenarios requiring a high picking accuracy.	Such as picking randomly placed sheet metal parts, shafts, track links, inner wheels, rings, and brake discs, and loading neatly arranged crankshafts, camshafts, and shaft yokes
Use a camera to acquire the point cloud, generate the target object model, and configure the pick point manually	Get point cloud by camera	There is no STL model or poor consistency between the acquired point cloud and the STL model file. There is no noticeable change in the surface feature point cloud of the target object. Single incoming orientation of target objects. Scenarios with low requirements for picking accuracy.	Bin recognition
Generate point cloud model based on common 3D shapes, and configure pick point manually	Create common 3D shape	The shapes of the target objects to be recognized are relatively regular, matching cylinder or cuboid.	Such as picking neatly arranged shafts, rings, and rectangular target objects, and picking randomly stacked shafts
Set the pick points by jogging the robot, and generate the point cloud model based on the acquired point cloud	Jog robot and get point cloud	Scenarios requiring a high picking accuracy.	Engine housing picking, compressor depalletizing
Import a processed point cloud to generate the point cloud model and pick points	Import processed point cloud	Scenarios with special requirements for preprocessing of point cloud models.	Scenarios requiring point cloud stitching
Target object configuration workflow when no point cloud model is required	No point cloud model required	Depalletizing; loading and unloading scenarios without using 3D matching methods for pallets with partitions or bins; loading and unloading of regular geometric target objects; generating point cloud models based on known target object dimensions.	Carton and sack depalletizing, brake disc loading and unloading, and scenarios for production increase based on target object dimensions

Basic operation procedure

Configuration workflow displayed in the software

Application scenario

Example

Import STL file to generate point cloud model and configure pick points manually

Import STL model

The consistency between the point cloud acquired by the camera and the STL model file is high.
The diversity of incoming target object orientations, along with the changes in the target object pose, will lead to significant changes in the features on the target object’s point cloud.
The surfaces of the target objects are reflective, and the missing areas in the point clouds vary between different objects.
Scenarios for picking neatly arranged objects.
Scenarios requiring a high picking accuracy.

Such as picking randomly placed sheet metal parts, shafts, track links, inner wheels, rings, and brake discs, and loading neatly arranged crankshafts, camshafts, and shaft yokes

Use a camera to acquire the point cloud, generate the target object model, and configure the pick point manually

Get point cloud by camera

There is no STL model or poor consistency between the acquired point cloud and the STL model file.
There is no noticeable change in the surface feature point cloud of the target object.
Single incoming orientation of target objects.
Scenarios with low requirements for picking accuracy.

Bin recognition

Generate point cloud model based on common 3D shapes, and configure pick point manually

Create common 3D shape

The shapes of the target objects to be recognized are relatively regular, matching cylinder or cuboid.

Such as picking neatly arranged shafts, rings, and rectangular target objects, and picking randomly stacked shafts

Set the pick points by jogging the robot, and generate the point cloud model based on the acquired point cloud

Jog robot and get point cloud

Scenarios requiring a high picking accuracy.

Engine housing picking, compressor depalletizing

Import a processed point cloud to generate the point cloud model and pick points

Import processed point cloud

Scenarios with special requirements for preprocessing of point cloud models.

Scenarios requiring point cloud stitching

Target object configuration workflow when no point cloud model is required

No point cloud model required

Depalletizing; loading and unloading scenarios without using 3D matching methods for pallets with partitions or bins; loading and unloading of regular geometric target objects; generating point cloud models based on known target object dimensions.

Carton and sack depalletizing, brake disc loading and unloading, and scenarios for production increase based on target object dimensions

Is this page helpful?

Leave a feedback

Select Configuration Workflow for Target Objects

Is this page helpful?

Thanks for your support!

You can give a feedback in any of the following ways:

We Value Your Privacy