Robot-Mounted 3D Measurement Integration for Faster Registration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing algorithms for 3D data registration in factory automation, such as those described in Non-Patent Literatures 1 and 2, are computationally intensive and lack robustness, particularly for objects with complex shapes or large displacements, leading to lengthy processing times that hinder productivity.
Innovation Solution
A measurement system comprising a 3D sensor, a displacement detector, a drive, a sensor controller, an integrator, and a robot controller, which performs 3D integration of data before and after a robot operation, allowing for more robust and efficient generation of 3D data, even during continuous robot operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If model-based algorithms (Non-Patent Literatures 1 and 2) are used for 3D data registration, then measurement precision is improved, but computing time increases significantly and robustness decreases
Solution Approach 1:
The patent applies preliminary action by performing coarse registration using model-based algorithms before fine registration with ICP. This preliminary coarse registration establishes an initial transformation that brings points closer together, reducing the computational burden and iteration requirements for subsequent fine registration, thereby decreasing total computing time while maintaining accuracy
Solution Approach 2:
The patent segments the registration process into two distinct stages: coarse registration (preprocessing) and fine registration (postprocessing). This segmentation allows each stage to use the most appropriate algorithm for its specific purpose, optimizing both speed and accuracy without requiring the entire process to use computationally intensive methods
2Measurement precision
If ICP algorithm is used for fine registration, then measurement precision is improved, but the algorithm may fail or require numerous iterations for points with large displacements, increasing computing time
Solution Approach 1:
The coarse registration performed before ICP serves as a preliminary action that reduces large displacements between corresponding points. By establishing an initial transformation that brings points closer together, the subsequent ICP fine registration operates on points with smaller discrepancies, ensuring convergence and reducing iteration requirements, thereby improving both reliability and computing time
3Measurement precision
If multiple measurement points are used to measure complex shapes or mirror-finished surfaces, then measurement precision is improved, but the complexity of the measurement system increases
Solution Approach 1:
The patent applies universality by using the same range sensor for both measurement and for providing robot positioning information through displacement detection. The sensor serves multiple functions: capturing 3D point cloud data for registration and providing displacement data for coordinate transformation, thereby reducing the need for separate measurement systems and reducing overall system complexity
Data Source
AI summary
A measurement system and associated techniques can reduce the time for obtaining 3D data about a target object by measurement. The measurement system includes a 3D sensor installable on a robot to obtain, by measurement, 3D data indicating 3D coordinates of surface points on a surface of a target object, a displacement detector that detects a displacement of a joint of the robot, a drive that drives the joint of the robot, a sensor controller that controls the 3D sensor to obtain 3D data about the target object by measurement at one or more measurement points, an integrator that performs 3D integration of 3D data about the target object obtained by measurement before a first operation performed by the robot on the target object and 3D data about the target object obtained by measurement at the one or more measurement points after the first operation performed by the robot on the target object, and a robot controller that outputs, to the drive, a drive command for controlling a second operation performed by the robot based on 3D coordinates of the surface points on the surface of the target object indicated by 3D data obtained through the 3D integration.