3D Object Orientation Measurement Using Transformation Parameters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for measuring the three-dimensional position and orientation of objects, particularly those with symmetrical or confusable front and back sides, often result in erroneous recognition due to the convergence of orientations during the fitting process, making it difficult to accurately determine the correct position and orientation.
Innovation Solution
An information processing apparatus that includes a model acquiring unit, a display control unit, a setting unit, a parameter deriving unit, and a storage unit to register and utilize transformation parameters between confusable positions and orientations, allowing for accurate calculation of the target object's position and orientation by generating and comparing fitting results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If model fitting method is used to measure three-dimensional position and orientation, then measurement capability is improved, but erroneous recognition occurs for objects with confusable orientations
Solution Approach 1:
The invention performs preliminary detection to obtain an initial position and orientation before model fitting. This initial value guides the subsequent model fitting process, preventing convergence to incorrect solutions for objects with confusable orientations like front and back sides that appear similar.
Solution Approach 2:
The invention uses the detection result as feedback to generate an initial value for model fitting, and then uses the model fitting result to refine the position and orientation measurement. This feedback loop ensures that the measurement process leverages information from both detection stages to achieve accurate results.
2Productivity
If detection method is used to determine initial position and orientation, then processing speed is improved, but incorrect solutions are obtained for symmetrical objects
Solution Approach 1:
The detection method is used as a preliminary step to quickly obtain an initial position and orientation estimate. This initial estimate is then refined through model fitting, combining the speed advantage of detection with the accuracy advantage of model fitting.
Solution Approach 2:
The invention changes the parameter representation by using the detection result as an initial value parameter for the model fitting process. This parameter transformation allows the system to leverage the fast detection result while ensuring convergence to the correct solution through subsequent model fitting with proper initial conditions.
Data Source
AI summary
A technique is provided, with which, even for an object which has a shape prone to be erroneously recognized and for which it is difficult to generate an initial value expected to produce a true value through fitting, it is possible to generate an appropriate initial value candidate through intuitive input and reduce erroneous recognition.An information processing apparatus includes a model acquiring unit configured to acquire a three-dimensional shape model of a target object, a display control unit configured to cause a display unit to display the three-dimensional shape model, a setting unit configured to set first and second orientations different from each other on the basis of the three-dimensional shape model displayed in the display unit, a parameter deriving unit configured to derive a transformation parameter that transforms the first orientation into the second orientation, and a storage unit configured to store the transformation parameter.