3D Object Identification for Stacked Components on Transparent Trays
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Two-dimensional machine vision in robotics for customized production often misidentifies complex components due to stacking, leading to incorrect selection by mechanical arms, especially when components are on transparent trays.
Innovation Solution
An object identification system utilizing a three-dimensional camera with a mechanical arm, which captures three-dimensional images and employs point cloud computing, depth image generation, and layer separation to accurately identify and grasp components by adjusting the camera position and angle, and calculating grasping points based on maximum planes and curvature values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If two-dimensional machine vision is used for component identification, then the system structure remains simple, but identification accuracy deteriorates due to stacking components and transparent trays
Solution Approach 1:
The patent transitions from two-dimensional machine vision to three-dimensional vision systems (using stereo cameras, time-of-flight cameras, or structured light scanners) to capture depth information. This dimensional upgrade enables the system to distinguish stacked components by their spatial positions and penetrate transparent tray backgrounds, resolving the identification accuracy problem while accepting increased system complexity.
Solution Approach 2:
The patent introduces point cloud processing and depth image generation as intermediary steps between raw image capture and component identification. These intermediate representations transform complex 3D visual data into structured spatial information, enabling accurate identification of stacked components while managing the complexity of 3D vision processing.
2Measurement precision
If three-dimensional imaging and processing techniques are used to accurately identify stacked components, then identification accuracy improves, but system complexity and processing time increase
Solution Approach 1:
The patent segments the 3D vision processing into distinct modules: depth image generation, point cloud processing, layer separation, and component identification. This modular segmentation manages system complexity by organizing 3D processing tasks into independent, manageable stages, each handling specific aspects of the identification problem.
Solution Approach 2:
The patent performs preliminary actions by generating depth images and point cloud representations before actual component identification. These pre-processed 3D data structures organize spatial information in advance, making the subsequent identification process more efficient and manageable despite the initial processing overhead.
3Measurement precision
If three-dimensional imaging is used to distinguish stacked components, then identification accuracy improves, but processing time increases
Solution Approach 1:
The patent performs preliminary depth image generation and point cloud processing to organize 3D spatial information before identification. By pre-structuring the spatial data, the system reduces processing time during actual identification tasks, as the heavy 3D computation is done once rather than repeatedly for each identification query.
Solution Approach 2:
The patent segments processing tasks to identify only the necessary 3D features for component differentiation. By focusing computation on critical spatial parameters (depth, layer position, component geometry) rather than processing all 3D data equally, the system achieves accurate identification with reduced processing time.
Data Source
AI summary
A method for identifying objects by shape in close proximity to other objects of different shapes obtains point cloud information of multiple objects. The objects are arranged in at least two trays and the trays are stacked. A depth image of the objects is obtained according to the point cloud information, and the depth image of the objects is separated and layered to obtain a layer information of all the objects. An object identification system also disclosed. Three-dimensional machine vision is utilized in identifying the objects, improving the accuracy of object identification, and enabling the mechanical arm to accurately grasp the required object.