Optical Surface Measurement via AI Phase Retrieval From Collimated Beams
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing optical metrology methods for measuring optical surfaces using divergent beams face challenges in accurately determining surface profiles without relying on wavefront reconstruction, particularly in setups where the relationship between the beam wavefront and the surface is complex and non-linear, such as in the Ritchey-Common test.
Innovation Solution
A method utilizing AI image space phase retrieval to directly measure optical surfaces by collimating a divergent beam, capturing a convergent beam image, and employing an artificial intelligence model to estimate the surface profile through a non-linear mapping function, without the need for wavefront reconstruction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If wavefront reconstruction is used to measure optical surfaces with divergent beams, then measurement capability is achieved, but measurement precision deteriorates due to complex non-linear mapping and data reduction requirements
Solution Approach 1:
The patent extracts the phase information directly from the intensity distribution in the image space, eliminating the need for complex wavefront reconstruction. By using phase retrieval algorithms that work directly with intensity measurements, the method removes the intermediate wavefront representation step that introduces complexity and potential errors in traditional interferometric approaches.
Solution Approach 2:
The patent replaces the traditional mechanical/optical wavefront sensing approach with an AI-based image processing system. Instead of using complex optical components and mechanical data reduction techniques, the invention uses machine learning models trained to directly map image intensity patterns to surface profiles, substituting physical measurement complexity with computational intelligence.
2Reliability
If traditional interferometric methods are used in Ritchey-Common test setup, then surface measurement is possible, but reliability deteriorates due to sensitivity to vibrations and alignment requirements
Solution Approach 1:
The patent creates a digital copy of the wavefront phase information from the intensity image through phase retrieval algorithms. This digital representation can be processed and analyzed without requiring the physical optical interference pattern to remain stable, making the measurement process more robust against vibrations and alignment drift that would disrupt traditional interferometric fringes.
Solution Approach 2:
The patent performs preliminary phase retrieval from the intensity image before any vibration or alignment changes can affect the measurement. By extracting phase information directly from the captured image through computational algorithms, the system establishes the surface profile data before environmental disturbances can degrade the optical path stability required by traditional interferometry.
3Productivity
If complex data reduction techniques are applied to RC test interferograms, then surface profile can be retrieved, but productivity deteriorates due to time-consuming processing
Solution Approach 1:
The patent applies partial phase retrieval by focusing on extracting only the essential surface profile information from the intensity image, rather than reconstructing the complete wavefront field. This selective approach retrieves sufficient phase data to determine surface shapes and deformations without performing exhaustive wavefront analysis, significantly reducing processing time while maintaining measurement accuracy for the intended application.
Solution Approach 2:
The patent changes the fundamental parameter being measured from wavefront phase (which requires complex interferometric processing) to intensity distribution in image space (which can be processed directly). By formulating the measurement problem in terms of intensity patterns that can be analyzed through phase retrieval algorithms, the system transforms a computationally intensive wavefront reconstruction problem into a more efficient image processing task.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables accurate and efficient measurement of optical surfaces by directly retrieving surface profiles from captured images, improving signal-to-noise ratio and robustness against vibrations, and eliminating the need for complex data reduction techniques.
Implementation Method 1
at least one referenceable collimator providing a measurement reference for the tested surface and collimating a beam travelling through the optical system
Implementation Method 2
reflecting the convergent beam from the referenceable collimator, to the testing surface, and to the beam transceiver
Data Source
AI summary
A method implemented by an optical system and a computing system allows for measuring of an optical surface by collimating a divergent beam. A divergent beam is emitted from the beam transceiver, to the testing surface, and to the referenceable collimator in order to convert the divergent beam into a convergent beam. The convergent beam is then reflected from the referenceable collimator, to the testing surface, and to the beam transceiver. The beam transceiver then captures a current image of the convergent beam. The computing system then executes an estimation process by inputting the current image into the estimation process. The computing system finally executes the estimation process with the computing system by outputting a surface profile of the testing surface with the estimation process.