Frame Transformation

You are currently viewing the documentation for a pre-release version (2.2.0). To access documentation for other versions, click the "Switch Version" button located in the upper-right corner of the page.

■ If you're unsure about the version of the product you are using, please contact Mech-Mind Technical Support for assistance.

This section describes the Frame Transformation function in the Advanced Pose Transformation step and how to use it.

Function Introduction

This function transforms input poses or point clouds from the current reference frame to a custom reference frame. By providing a reference pose that defines the custom frame, data can be aligned and switched between different frames.

Workflow

The following sections describe how to use this function for pose frame transformation and point-cloud frame transformation.

Transform Pose Reference Frame

Before transforming pose reference frames, first configure step ports.

Configure Step Ports

  1. Click Configuration Wizard to open the pose-transformation tool, and select Frame Transformation.

  2. In the parameter panel on the right, set transformation type to Pose.

  3. Click OK to complete configuration.

Configured input and output ports are as follows:

Input

Input Port Data Type Description

Raw Pose

Pose []

Input poses at this port are transformed to another frame.

Custom Reference Frame

Pose []

Input poses are transformed to the custom frame defined by this port.

Output

Output Port Data Type Description

Transformed Pose

Pose []

Pose data after frame transformation.

The pose to be transformed and the custom reference pose must be in the same coordinate system.

Transform Pose Reference Frame

By inputting Raw Pose and Custom Reference Frame, raw poses can be transformed into the custom frame. The following describes how to use the Advanced Pose Transformation step for pose-frame transformation with typical project scenarios.

Scenario 1: Sort Workpieces by Bin-Z When Bin Is Tilted

In production, bins may be globally tilted, so Z values in camera frame cannot reflect true height order inside the bin. In this case, workpiece poses should be transformed to the bin frame and sorted along the bin Z axis for in-bin sorting or layered picking.

  1. Use workpiece poses in camera frame as Raw Pose, and use detected bin pose as Custom Reference Frame. Use Advanced Pose Transformation to transform workpiece poses to bin frame.

  2. In bin frame, use pose-sorting steps to sort workpieces by Z axis.

transform pose 1

After sorting workpiece poses, you usually need to convert sorted poses back to the original frame. This can be done by Transform Poses. Use Sorted Poses as Raw Pose, and input Custom Reference Frame from Advanced Pose Transformation as Reference Pose, then the inverse transformation from custom frame to original frame is performed.
Scenario 2: Compute Relative Height When Reference Plane Is Tilted

In some scenarios, workpieces are placed on a non-horizontal reference plane (for example, tilted conveyor or fixture). Directly reading Z values in camera frame cannot accurately reflect real height relative to the reference plane.

To obtain relative height, transform workpiece poses to a custom frame whose XY plane matches the reference plane and whose Z axis is perpendicular to it, then compute relative height accurately.

  1. Use steps such as Calculate Poses and Dimensions from Planar Point Clouds to compute a reference pose (custom frame) from reference-plane point cloud in camera frame.

  2. Use Advanced Pose Transformation to transform raw workpiece poses to this custom frame.

  3. Decompose transformed poses and extract Z component from translation vector as workpiece height relative to reference plane.

transform pose 2

The transformed Z value is the actual workpiece height relative to the reference plane. If the pose is needed later in original frame, restore it using Transform Poses.

Transform Point-Cloud Reference Frame

Before transforming point-cloud reference frames, first configure step ports.

Configure Step Ports

  1. Click Configuration Wizard to open the pose-transformation tool, and select Frame Transformation.

  2. In the parameter panel on the right, set transformation type to Point Cloud.

  3. Click OK to complete configuration.

Configured input and output ports are as follows:

Input

Input Port Data Type Description

Raw Point Cloud

PointCloud []

Input point cloud at this port is transformed to another frame.

Custom Reference Frame

Pose []

Input point cloud is transformed to the custom frame defined by this port.

Output

Output Port Data Type Description

Transformed Point Cloud

PointCloud []

Point cloud after frame transformation.

The point cloud to be transformed and the custom reference pose must be in the same coordinate system.

Transform Point-Cloud Reference Frame

By inputting Raw Point Cloud and Custom Reference Frame, the raw point cloud can be transformed to the custom frame. The following describes how to use Advanced Pose Transformation for point-cloud frame transformation with typical project scenarios.

Scenario 1: Camera View Direction Is Not Consistent with Workpiece Plane

In real applications, due to factors such as workpiece reflection, a 3D camera often cannot be mounted perpendicular to workpiece surface and is installed at a tilt angle. As a result, captured point cloud appears tilted in camera frame. If point filtering is performed directly in this frame, valid points may be removed and noise points may be retained, affecting plane fitting, feature extraction, and measurement accuracy.

To improve processing quality, build a custom frame from main workpiece-plane point cloud (Z axis perpendicular to workpiece plane), transform raw point cloud from camera frame to workpiece local frame by Advanced Pose Transformation, preprocess point cloud there, then restore point cloud to camera frame.

  1. Use steps such as Calculate Poses and Dimensions from Planar Point Clouds to compute a reference pose (custom frame) from workpiece-plane point cloud in camera frame.

  2. Use Advanced Pose Transformation to transform raw workpiece point cloud into this custom frame for point-cloud processing.

transform point cloud 1

After point-cloud preprocessing, you usually need to restore processed point cloud to original frame. In Transform Point Clouds, select UseCorrespondenceInput transformation type, use Processed Point Cloud as Raw Point Cloud, and input Custom Reference Frame from Advanced Pose Transformation as Reference Pose to perform inverse transformation from workpiece frame to camera frame.
Scenario 2: Obtain Accurate Circle Center (Based on Orthographic Projection)

In some high-precision applications, circle-center positions and similar geometric features are hard to obtain accurately by direct 3D matching. To improve detection accuracy, combine point-cloud frame transformation and orthographic projection: transform point cloud to workpiece frame, extract 2D features, and then restore results to camera frame to obtain higher-precision circle-center poses. This method is commonly used for high-precision positioning of assembly holes or arc edges.

  1. Use steps such as Calculate Poses and Dimensions from Planar Point Clouds to compute a reference pose (custom frame) from workpiece-plane point cloud in camera frame.

  2. Use Advanced Pose Transformation to transform raw workpiece point cloud to this custom frame.

  3. For transformed point cloud, use steps such as Orthographic Projection for 2D feature extraction and circle-center pose calculation.

transform point cloud 2

After obtaining accurate circle-center poses, you usually need to restore them to camera frame. This can be done by Transform Poses: use Calculated Circle-Center Pose as Raw Pose, and input Custom Reference Frame from Advanced Pose Transformation as Reference Pose, then inverse transformation from workpiece frame to camera frame is performed.

Is this page helpful?

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

We Value Your Privacy

We use cookies to provide you with the best possible experience on our website. By continuing to use the site, you acknowledge that you agree to the use of cookies. If you decline, a single cookie will be used to ensure you're not tracked or remembered when you visit this website.

  • Preference
  • I accept
  • I don't accept