Non-Planar Screen 3D Measurement Without Calibration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing measurement methods for the three-dimensional shape of objects require calibration of apparatuses with unknown internal parameters, making the process time-consuming and inefficient.
Innovation Solution
A measurement method that generates Euclidean restoration information for a non-planar screen surface using known internal parameters, and then transforms projection restoration information from an apparatus with unknown internal parameters into Euclidean restoration information using a three-dimensional projective transformation matrix.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If calibration is performed to grasp internal parameters of measurement apparatus, then measurement precision is improved, but measurement time increases
Solution Approach 1:
The patent introduces a reference object with known three-dimensional shape as an intermediary between the measurement apparatus and the target object. By capturing images of this reference object and calculating internal parameters from the known geometry, the system obtains accurate calibration data without requiring time-consuming traditional calibration procedures. The reference object serves as a mediator that enables parameter extraction through geometric calculation rather than iterative calibration.
Solution Approach 2:
The measurement apparatus performs self-calibration by using the reference object to automatically calculate its own internal parameters. The system captures images of the reference object, uses the known three-dimensional coordinates to compute internal parameters through projective geometry, and then uses these parameters for subsequent measurements. This self-service approach eliminates the need for external calibration equipment or procedures.
2Measurement precision
If traditional calibration procedures are used to obtain internal parameters, then measurement accuracy is improved, but device complexity and operation difficulty increase
Solution Approach 1:
The system automatically calculates internal parameters by capturing images of the reference object and using the known three-dimensional coordinates to compute parameters through projective geometry. This automated self-service process eliminates manual calibration operations, reducing both device complexity and operational difficulty while maintaining high measurement accuracy.
Solution Approach 2:
The patent uses a digital model (copy) of the reference object with known three-dimensional coordinates instead of physical calibration artifacts. By working with the digital geometric representation, the system simplifies the calibration process to image capture and calculation, eliminating the need for complex physical calibration setups and procedures.
Data Source
AI summary
A measurement method including generating first restoration information representing a Euclidean restoration result for a first region of a screen that is a non-planar surface based on a first image that is one of an image input to a first apparatus including a first lens and having known internal parameters relating to the first lens and an image output from the first apparatus, and a second image that is one of an image input to a second apparatus including a second lens and having known internal parameters relating to the second lens and an image output from the second apparatus, generating second restoration information representing a projection restoration result for a second region of the screen based on a third image that is one of an image input to a third apparatus including a third lens and having unknown internal parameters relating to the third lens and an image output from the third apparatus, and the first image, generating a three-dimensional projective transformation matrix for transforming the coordinates of a plurality of points indicated by the second restoration information into the coordinates of a plurality of points indicated by the first restoration information based on the first restoration information and the second restoration information, and generating third restoration information representing a Euclidean restoration result for the second region based on the second restoration information and the three-dimensional projective transformation matrix, at least one of the first, second, and third apparatuses being an imaging apparatus that captures an image of the screen, and at least part of the first region and at least part of the second region overlapping with each other in a third region.