2D-3D Marker Detection With Plane Projection for Robot Teaching

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing marker detection systems, such as those using 2D cameras, suffer from position deviation and distortion due to lens distortion, leading to inaccurate marker detection, which affects the precision of robot teaching systems.

Innovation Solution

A marker detection apparatus that combines 2D and 3D camera data to generate a combined image, using point group data to accurately determine marker planes and correct marker positions, ensuring high-precision detection.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If a 2D camera is used to detect the marker, then the device complexity is reduced, but the measurement precision deteriorates due to lens distortion and viewing angle effects

Engineering Contradiction:
Improvecamera systemVSAvoidmarker position
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The patent combines 2D camera image data with 3D camera point group data to create a combined image. The 2D camera provides high-resolution marker identification while the 3D camera provides depth information and spatial context. By merging these data sources, the system achieves accurate marker position detection without relying solely on the distorted 2D camera images.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent introduces a marker plane as an intermediary concept. The marker plane is determined based on point group data from the 3D camera, and the marker position is projected onto this plane to correct for viewing angle and distortion effects. This intermediary plane serves as a reference that bridges the 2D image data and 3D spatial information.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Loss of information

If 3D camera point group data alone is used for marker detection, then depth information is improved, but the measurement precision deteriorates due to low resolution

Engineering Contradiction:
Improvedepth informationVSAvoidmarker position
Core Design Contradiction:
Loss of informationVSMeasurement precision

Solution Approach 1:

The patent merges the high-resolution 2D camera image data with the depth-rich 3D camera point group data. The 2D camera provides precise marker contour and position information, while the 3D camera provides accurate depth and spatial plane information. The combination allows the system to overcome the low resolution limitation of the 3D camera by using the 2D camera's high-resolution image data for final position determination.

Inventive Principle:
Principle #5Merging (Combining)

3Device complexity

If marker detection is performed without plane determination, then the device complexity is reduced, but the measurement precision deteriorates due to contour deviation and distortion

Engineering Contradiction:
Improveprocessing stepsVSAvoidmarker position
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The patent performs preliminary plane determination based on point group data from the 3D camera before detecting the marker position. By establishing the marker plane in advance using the reliable depth information, the system creates a reference framework that guides the subsequent marker detection process. This preliminary action allows the system to correct for distortion and viewing angle effects during the detection phase.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The marker plane acts as an intermediary reference that mediates between the raw 3D point cloud data and the final 2D marker position detection. The plane is determined from 3D data and then used to constrain and guide the 2D marker detection, ensuring that the detected position is projected onto the correct spatial plane and corrected for perspective and distortion effects.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentUS12397437B2Marker detection apparatus and robot teaching system
Publication Date: 2025.08.26 DAIHEN CORP
  • US12397437B2 patent drawing
  • US12397437B2 patent drawing
  • US12397437B2 patent drawing

AI summary

Provided is a marker detection apparatus able to detect the position of a marker with high accuracy and a robot teaching system using the same. A marker detection apparatus includes: an image generation unit that superimposes point group data, acquired by a 3D camera and photographed from a viewpoint same as a viewpoint of a 2D camera, on image data acquired by the 2D camera, thereby generating a combined image; a marker position calculation unit that detects a marker and calculates a position of the marker on a basis of the image data; a plane detection unit that detects a plane on a basis of the point group data; a marker plane determination unit that determines a marker plane on a basis of point group data included in a region of the marker detected from the position of the marker calculated by the marker position calculation unit and the plane detected by the plane detection unit; and a marker position correction unit that projects the position of the marker, calculated by the marker position calculation unit, onto the marker plane, thereby correcting the position of the marker.