Real-Time Colorimetry Compensation for Non-Uniform Projection Surfaces
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods fail to effectively compensate for colorimetry issues in projected images on non-uniform surfaces, especially in the presence of external light, due to the complexity of determining the necessary kernel values and the limitations of projector power.
Innovation Solution
A real-time colorimetric compensation method that simplifies the projection model by using a target image with global changes and expressing the kernel as a linear combination of bases, reducing the number of parameters to be determined, and accounting for the low-frequency nature of texture and external light.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the full projection model with complete kernel determination is used, then measurement precision is improved, but device complexity increases significantly
Solution Approach 1:
The patent segments the projection model into distinct components: the transfer function P representing projector characteristics, and the kernel K representing projection surface properties. By separating these components and determining them independently through separate calibration processes, the system avoids the complexity of determining the complete kernel while maintaining colorimetry compensation accuracy.
Solution Approach 2:
The patent extracts and removes the transfer function P from the full projection model, determining it separately through calibration with a uniform reference surface. This extraction simplifies the remaining problem to determining only the kernel K, significantly reducing device complexity while preserving measurement precision for colorimetry compensation.
2Productivity
If real-time compensation is implemented, then productivity is improved, but measurement precision may deteriorate due to simplified models
Solution Approach 1:
The patent performs preliminary calibration actions before real-time operation: the transfer function P is determined through calibration with a uniform reference surface, and the kernel K is determined through calibration with a color chart. These preliminary determinations store the projection characteristics, enabling real-time compensation without complex calculations during actual projection, thus maintaining both speed and precision.
Solution Approach 2:
The patent implements a feedback mechanism where the determined transfer function and kernel are applied to compensate projected images in real-time. The system continuously uses the calibrated parameters to adjust and correct colorimetry issues as images are projected, ensuring ongoing precision while maintaining real-time processing capability.
3Measurement precision
If the complete kernel is determined, then measurement precision is improved, but loss of time increases due to huge number of values
Solution Approach 1:
The patent segments the kernel determination process into manageable calibration steps using standardized reference surfaces. By determining the transfer function first with a uniform surface and then the kernel with a color chart, the system avoids the need to determine all kernel values simultaneously, significantly reducing the time required while maintaining measurement precision.
Solution Approach 2:
The patent changes the parameters being measured during calibration by using different reference surfaces with known properties (uniform surface for transfer function, color chart for kernel). This approach allows efficient determination of projection characteristics without requiring exhaustive measurement of all possible kernel values, reducing time loss while preserving accuracy.
Data Source
Figure 1~2
Figure 3~4
Figure 5
AI summary
Real time compensation of the colorimetry of an image projected on a nonuniform projection surface is disclosed. The projected image is captured and coefficients of a simplified model of the projection surface are estimated by minimizing the distance between the captured image and a target image obtained from the original image.