The project is the core work unit of PickWiz, including Robot configuration and communication, Eye-hand calibration, visual Parameter and other functions and Task configurations. It provides an efficient and flexible solution for Robot to complete picking tasks, and meets the needs of many types of industrial scenes such as depalletizing sacks and cartons, disorderly grabbing, orderly loading and unloading, positioning and assembly, etc.

The process of building the project is as follows:

1. Create a new project

When you open the PickWiz software for the first time, you will enter the welcome page. If you need to create a new empty project, please refer to 1.1 Create a new empty project;

For common typical application Scene, you can load the required Scene project from Case Library, which contains the basic configuration data of the Scene, which can debug the configuration more quickly. For details, please refer to [1.2 Create a new project from the case library] (#12-从案例库新建项目).

1.1 Create a new empty project

Select TaskScene and press Scene to create a new empty project

①Select TaskScene

Click the mouse on the Scene image to display the TaskScene introduction.

If the materials are sacks and cartons, please check Depalletizing Scene

If the materials are stacked and disorderly arranged Targets, please refer to Disordered Fetching Scene

If the materials are stacked and orderedly placed Targets, please refer to Orderly Loading and Unloading Scene

If there is only a single Target that needs to be positioned within the camera's field of view, please refer to [Positioning Assembly] (#dingwei)

Scene 1: Destacking Scene

Depalletizing refers to the process of separating and transporting stacked materials to specific locations one by one.

• If the material is a single type of sack, select Single Unpacking of Sacks

• If the material is a carton, select Carton Single Disassembly

Scene 2: Fetching Scene out of order

The disordered grabbing Scene refers to the robot identifying and grabbing targets that are arranged in an orderly and irregular manner to complete the loading and unloading tasks.

• If the Target is round and arranged randomly and irregularly, select Picking of randomly arranged Circle type target objects

• If the Target is a cylinder and is arranged randomly and irregularly, select Cylinder-based disordered grabbing

• If the Target is not round or cylindrical, has a large difference between the front and back sides, and is disorderly and irregularly placed, select Picking randomly arranged general type target objects. This Scene is suitable for most disordered grabbing Scenes, but it is not applicable to Targets with small differences between the front and back sides.

• If the target is not round or cylindrical, the difference between the front and back sides is small, and the target is disorderly and irregularly placed, then select Unordered grabbing of surface-type Target

• If the Target is not round or cylindrical, the difference between the front and back sides is small, the Targets are isolated from each other and arranged disorderly and irregularly, then select Surface Target loading and unloading (materials are isolated from each other)

Scene 3: Orderly loading and unloading Scene

The orderly loading and unloading scene refers to the robot identifying and grabbing the targets placed in an orderly and regular manner to complete the task of loading and unloading materials.

• If the Target is round and arranged regularly and orderly, select Loading and unloading of Circle type target objects

• If the Target is a cylinder and is placed in a regular and orderly manner, select Ordered loading and unloading based on cylinder

• If the Target is quadrilateral and placed in a regular and orderly manner, select Loading and unloading of Quadrilateral type target objects

• If the Target is not a cylindrical quadrilateral with a large difference between the front and back sides and is arranged in a regular and orderly manner, select Loading and unloading of neatly arranged general type target objects. This Scene is suitable for most ordered loading and unloading scenes, but it is not applicable to Targets with small differences between the front and back sides.

• If the target is not a round cylindrical quadrilateral, the difference between the front and back sides is small, and the target is arranged in a regular and orderly manner, then select Organized loading and unloading of surface-type Target

• If the target is not a cylindrical quadrilateral with a round surface, the difference between the front and back sides is small, and the front and back sides are arranged in a regular and orderly manner, and it is also necessary to significantly reduce the visual calculation time, then choose Ordered loading and unloading of surface-type Target (parallelization)

• If the target is not a cylindrical quadrilateral with a small difference between the front and back sides and is arranged in a regular and orderly manner, and rapid deployment is required, select Loading and unloading of neatly arranged plane type target objects (image matching)

Scene four: Positioning assembly

Positioning assembly means that there is only a single Target that needs to be positioned within the camera's field of view, such as single-target precise positioning of the charging head so that the robotic arm can accurately insert and remove the charging head.

• If there is only a single Target that needs to be positioned within the camera's field of view, the difference between the front and back of the Target is small, and the Target position changes little (ROI 3D can completely frame the Target, please see [ROI Operation Guide] (ROI操作指南.md) for details about ROI 3D), then select Plane type Target positioned assembly (only Point Cloud matches)

• If there is only a single Target that needs to be positioned within the camera's field of view, the difference between the front and back of the Target is small, and the Target position does not change much (ROI 3D can completely frame the Target, please see [ROI Operation Guide] (ROI操作指南.md) for details about ROI 3D), then select Plane type Target positioned assembly (image matching only)

• If there is only a single Target that needs to be positioned within the camera's field of view, the difference between the front and back of the Target is small, and the Target position changes greatly (ROI 3D cannot completely frame the Target, please see [ROI Operation Guide] (ROI操作指南.md) for details about ROI 3D), then select Plane type Target positioned assembly

• If there is only a single Target that needs to be positioned within the camera's field of view and the difference between the front and back of the Target is large, select General Target positioned assembly. This is not applicable to Targets with small differences between the front and back.

②After selecting TaskScene, fill in the project name. The default is "new_project" + project serial number, which can be modified by yourself.

The project name cannot contain Chinese characters or spaces.

③Select Project Path, the default project storage path is C:\Users\dex\Documents\PickWiz，可自行修改

④Check the function options

Each Scene has multiple Task types, and each Task type corresponds to multiple function options, which can be selected according to actual working conditions.

Function options	Description
Instance Optimization	optimizes the model generated instance and processes the instance mask.
Pallet Detection	is used to detect the bottom bracket. After checking, the visual Parameter interface will add Pallet Detection related Parameters. For the specific process, please refer to the Pallet Detection User Guide
Partition and Pallet Detection	is suitable for scenes where there are partitions between each layer of materials and a bottom bracket. The three data types of partitions, bottom brackets, and targets can be entered separately through visual classification. For the specific process, please refer to the Partition and Pallet Detection User Guide
Collision detection	is used to detect the collision between the End-effector and the container, and filter the grabbing gestures that may collide; after checking, you need to configure the container-related Parameters in "Scene Object", and the visual Parameter interface will add collision detection-related Parameters. For the specific process, please refer to the Collision Detection Usage Guide.
Type Identification	uses a deep learning model to identify multiple types of Targets; after checking, you need to configure multiple types of Targets according to your needs on the Target interface, and then use the built-in one-click connection training model. For specific procedures, please refer to the Type Identification User Guide
Local Identification	locates the Target through its local characteristics; after checking, you need to configure the overall information and local information of the Target according to your needs on the Target interface, and then use the built-in one-click connection training model. For specific procedures, please refer to Local Identification User Guide
Front-Back Identification (via Point Cloud template)	can import the Point Cloud templates of the front and back sides of the Target to match and capture the front or back sides of the Target, and the Pick Point can be set separately for the front or back sides of the Target. For the specific process, please refer to Front-Back Identification (via Point Cloud template) User Guide
Visual classification	is used to identify different textures, orientations and other features of the same Target. After turning it on, different types of characteristics of the same object can be entered on the Target interface, and related parameters can be set on the visual Parameter interface. For specific procedures, please refer to the Visual Classification User Guide

1.2 Create a new project from the case library

• Switch to New from Case Library on the welcome page and select the required Scene project

• If you are using this case for the first time, you need to click the Download button to download the project configuration.

• After downloading the project configuration of this case, fill in the project name and project path.

2. Load project

If you need to open an existing project on the disk, click Load Project on the welcome page

Open the corresponding project folder, select the project, and click Select Folder. The default project storage path is C:\Users\dex\Documents\PickWiz。

If the following incompatibility issues occur when opening a project, please use a lower version of PickWiz software to open the project. If necessary, please contact Dexforce technical support personnel

---

3. Project configuration

Open the project configuration folder config to view the project configuration data, as shown in the figure below

Notice

It is not recommended to modify the project configuration folder, otherwise PickWiz may report abnormal errors and cannot be used normally.

when the current Task is running when the current Task is running when the current Task is running when the current Task is running when the current Task is running. when the current Task is running The root configuration file of the

serial number	Folder name	configuration name	Description
1	calibration	Eye-hand calibration (new version)	stores the Eye-hand calibration (new version) configuration
2	camera	Camera	stores the Camera and CameraParameter
3	end_effector	End-effector	stores the End-effector configuration
5	PickLightMidData	Visual configuration	stores the configuration corresponding to the visualization window
6	product	Target	stores the Target configuration of the current Task
7	robot	Robot	stores the Robot
8	roi	ROI	stores the ROI configuration when the current Task is running
9	roi2d	ROI 2D	stores the ROI 2D configuration of the current Task runtime
10	roi3d	ROI 3D	stores the ROI 3D configuration
11	scene_object	Scene Object	stores various objects in the actual Scene of the current Task, including material frames, bottom brackets, etc.
12	strategy	Visual Parameter	stores the visual Parameter configuration
13	dexforce_proj.json	Root Profile	project. The configurations corresponding to all tasks of the project are specified in this file

After creating a new project, Pickwiz will create a default Task, and the project configuration data can be compared with the Task information.

Label	Description
Task name	The name of the current Task, which can be modified
Task type	The type of the current Task, i.e. TaskScene, corresponds to the TaskScene selected when creating a new Task and cannot be modified
TaskID	Identifies the current Task. Robot can switch Tasks through this ID
Camera	Camera & CameraParameter when the current Task is running. To connect the Camera and adjust the Camera imaging, please refer to Camera Connection and Parameter Adjustment Guide
Robot	The Robot when the current Task is running. To establish communication with the Robot, please refer to Robot Configuration and Communication Operation Guide
Target	The Target configuration of the current Task is configured to the Task information after adding it in the Target interface. To add a new Target, please refer to Target Operation Guide
End-effector	End-effector configuration when the current Task is running. Configure it to the Task information after adding it in the End-effector interface. To add an End-effector, please refer to End-effector User Guide
Scene Object	Various objects in the current TaskScene, including material frames, bottom brackets, etc., check Collision Detection, Pallet Detection, Partition and Pallet Detection in Function Options before the Scene Object interface will appear. After adding a Scene Object, you need to configure it to the Task information. For the specific process, please refer to [Pallet Detection User Guide] (../功能选项/底托检测使用指南.md), Collision Detection User Guide, Partition and Pallet Detection User Guide
Eye-hand calibration	The Eye-hand calibration configuration when the current Task is running is configured to the Task information after adding it in the Eye-hand calibration interface. For the specific process, please refer to [Eye-hand calibration Operation Guide] (../手眼标定/手眼标定操作指南.md)
ROI	The ROI configuration when the current Task is running is configured to the Task information after adding it in the ROI interface. For the specific process, please refer to ROI Operation Guide
Visual Parameter	The visual Parameter configuration when the current Task is running is configured in the Visual Parameter interface. For the specific process, please refer to the Visual Parameter Adjustment Guide
Function options	Click "Edit" to select "Collision Detection", "Type Identification", "Local Identification", "Pallet Detection" and other working condition attributes. Select the working condition attributes related to the project. PickWiz will match the functional configuration of the adapted working condition based on the selected working condition attributes, that is, add the configuration related to the selected working condition in related configurations such as Target and Visual Parameter. For related functions, please refer to the Function Options User Guide

Previous page3D Vision Guidance Kit Hardware Setup

Next pageTask Operation Guide