# US 2021/0279858 A1 · Material testing of optical test pieces

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2021/0279858 A1
- Document ID: US20210279858A1
- Status: Published application
- Kind: A1
- Application number: 17/251,415
- Filed: 2019-06-06
- Published: 2021-09-09
- Earliest claimed: 2018-06-12
- Assignees: Carl Zeiss Jena GmbH
- Applicants: Carl Zeiss Jena GmbH
- Inventors: Lars Stoppe; Niklas Mevenkamp; Alexander Freytag
- Web page: https://patentgrep.com/patent/US20210279858A1

## CPC

- G06T 7/0004: {Industrial image inspection}
- G01M 11/0221: {by determining the optical axis or position of lenses}
- G01M 11/0264: {by using targets or reference patterns}
- G01M 11/0278: {Detecting defects of the object to be tested, e.g. scratches or dust (investigating the presence of flaws or contamination on materials by optical means G01N21/88)}
- G06T 5/50: using two or more images, e.g. averaging or subtraction
- G06T 5/90: Dynamic range modification of images or parts thereof
- G06T 7/70: Determining position or orientation of objects or cameras (camera calibration G06T7/80)
- G06T 2207/10152: Varying illumination
- G06T 2207/20081: Training; Learning
- G06T 2207/20084: Artificial neural networks [ANN]
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G01M: TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR

## Abstract

The invention relates to techniques for material testing of optical test pieces, for example of lenses. Angle-variable illumination, using a suitable illumination module, and/or angle-variable detection are carried out in order to create a digital contrast. The digital contrast can be, for example, a digital phase contrast. A defect detection algorithm for automated material testing based on a result image with digital contrast can be used. For example, an artificial neural network can be used.

## Claims

### Claim 1

A method for material testing of an optical test object using an optical imaging system, wherein the method comprises:
for each of at least one pose of the test object in relation to the optical imaging system: respectively actuating the optical imaging system to capture at least two images of the test object by means of at least one of angle-variable illumination and angle-variable detection,
for each of the at least one pose: respectively processing the corresponding at least two images to generate a result image with a digital contrast, and
on the basis of the at least one result image which was generated on the basis of the at least two images for the at least one pose: carrying out material testing of the test object.

### Claim 2 (depends on 1)

The method as claimed in claim 1,
wherein carrying out the material testing comprises the application of a defect detection algorithm for automatically detecting defects.

### Claim 3 (depends on 2)

The method as claimed in claim 2,
wherein, for a detected defect, the defect detection algorithm provides at least one of a corresponding segmentation of the respective result image and a corresponding classification of the respective defect in relation to predefined defect classes.

### Claim 4 (depends on 2)

The method as claimed in claim 2,
wherein, for a detected defect, the defect detection algorithm determines a corresponding influence of the respective defect on an optical effect of the test object.

### Claim 5 (depends on 2)

The method as claimed in claim 2,
wherein the defect detection algorithm determines an aggregated influence on the optical effect of the test object for a plurality of detected defects.

### Claim 6 (depends on 2)

The method as claimed in claim 2,
wherein the defect detection algorithm comprises a trained artificial neural network, which is configured to receive one or more of the at least one result image as input map.

### Claim 7 (depends on 6)

The method as claimed in claim 6,
wherein the artificial neural network is configured to provide an output map which indicates a position and a classification of the defects detected.

### Claim 8 (depends on 6)

The method as claimed in claim 6, furthermore comprising:
carrying out retraining of the artificial neural network on the basis of a result of the material testing.

### Claim 9 (depends on 2)

The method as claimed in claim 2,
wherein the defect detection algorithm comprises at least one of a threshold analysis of the digital contrast of the at least one result image and a statistical analysis of the digital contrast.

### Claim 10 (depends on 2)

The method as claimed in claim 2,
wherein the defect detection algorithm comprises an anomaly detector, which detects defects of a priori unknown defect classes.

### Claim 11 (depends on 2)

The method as claimed in claim 2,
wherein the detected defects are marked in at least one error map,
wherein the method furthermore comprises:
comparing the at least one error map with a reference error map and
determining a quality value of the material testing on the basis of the comparison.

### Claim 12 (depends on 11)

The method as claimed in claim 11,
wherein the at least one pose comprises a plurality of poses which are associated with a plurality of result images,
wherein the defect detection algorithm is respectively applied to each of the plurality of result images to obtain a respective error map,
wherein the method furthermore comprises:
registering defects marked in the error maps in a global coordinate system,
wherein the registered error maps are compared to the reference error map.

### Claim 13 (depends on 11)

The method as claimed in claim 11,
wherein the at least one pose comprises a plurality of poses which are associated with a plurality of result images,
wherein the method furthermore comprises:
registering the plurality of result images in a global coordinate system,
wherein the defect detection algorithm is coherently applied to the registered result images for the purposes of detecting 3D defects.

### Claim 14 (depends on 1)

The method as claimed in claim 1,
wherein carrying out the material testing comprises the application of an algorithm with machine learning,
wherein the algorithm with machine learning is configured to receive the at least one result image as input and to provide a quality value of the test object as output.

### Claim 15 (depends on 14)

The method as claimed in claim 14,
wherein the input of the algorithm with machine learning comprises a sequence of sliding windows, wherein the sliding windows of the sequence of sliding windows image different regions of the at least one result image,
wherein the output of the algorithm with machine learning is determined on the basis of a combination of results of a sequence of results of the algorithm with machine learning,
wherein the sequence of results corresponds to the sequence of sliding windows.

### Claim 16 (depends on 1)

The method as claimed in claim 1, furthermore comprising:
for each of the at least one pose: varying at least one imaging modality when capturing the at least two images,
wherein the at least one imaging modality is selected from the following group: color of the light used; polarization of the light used; magnitude of the at least one of angle-variable illumination and angle-variable detection; orientation of the at least one of angle-variable illumination and angle-variable detection.

### Claim 17 (depends on 1)

The method as claimed in claim 1,
wherein the digital contrast is selected from the group of: virtual dark field; digital phase gradient contrast; and digital phase contrast.

### Claim 18 (depends on 1)

The method as claimed in claim 1,
wherein the optical imaging system furthermore comprises a correction lens which has an optical effect that is complementary to an optical effect of the test object.

### Claim 19 (depends on 1)

The method as claimed in claim 1,
wherein the at least one pose comprises a plurality of poses,
wherein the plurality of poses define a z-stack and/or an xy-stack.

### Claim 20

A computing unit which is configured to:
actuate an optical imaging system in order to capture, by means of at least one of angle-variable illumination and angle-variable detection, at least two respective images of a test object for each of the at least one pose of said optical test object in relation to the imaging system,
for each of the at least one pose: respectively process the corresponding at least two images to generate a result image with a digital contrast, and
on the basis of the at least one result image which was generated on the basis of the at least two images for the at least one pose: carry out material testing of the test object.

### Claim 21 (depends on 20)

The computing unit as claimed in claim 20,
wherein the computing unit is configured to perform material testing of said optical test object using said optical imaging system, wherein:
for each of at least one pose of the test object in relation to the optical imaging system: respectively actuating the optical imaging system to capture-at least two images of the test object by means of at least one of angle-variable illumination and angle-variable detection,
for each of the at least one pose: respectively processing the corresponding at least two images to generate a result image with a digital contrast, and
on the basis of the at least one result image which was generated on the basis of the at least two images for the at least one pose: carrying out material testing of the test object.
