Systems, devices, and methods for providing feedback on and improving the accuracy of super-resolution imaging
Systems, methods, and computer-readable media for feedback on and improving the accuracy of super-resolution imaging. In some embodiments, a low resolution image of a specimen can be obtained using a low resolution objective of a microscopy inspection system. A super-resolution image of at least a portion of the specimen can be generated from the low resolution image of the specimen using a super-resolution image simulation. Subsequently, an accuracy assessment of the super-resolution image can be identified based on one or more degrees of equivalence between the super-resolution image and one or more actually scanned high resolution images of at least a portion of one or more related specimens identified using a simulated image classifier. Based on the accuracy assessment of the super-resolution image, it can be determined whether to further process the super-resolution image. The super-resolution image can be further processed if it is determined to further process the super-resolution image.
1. A method comprising:
obtaining, by a computing system, a training data set comprising a plurality of low resolution images of artifacts and an assigned suitability classification for each artifact;
generating, by the computing system, a detection model for determining whether a detected artifact in a low resolution image is suitable for super-resolution generation by:
inputting the training data set into a machine learning algorithm; and
learning, via the machine learning algorithm, to identify low resolution images of artifacts that are suitable for super-resolution generation by determining at least a context of each artifact;
obtaining, by the computing system, a target low resolution image of a specimen;
scanning, by the computing system, the target low resolution image to identify a target artifact in the specimen;
analyzing, by the computing system via the detection model, the target artifact to determine whether the target low resolution image is suitable for super-resolution imaging; and
upon determining, by the computing system via the detection model, that the target low resolution image is suitable for the super-resolution generation, performing super-resolution generation on the target low resolution image.
2. The method of claim 1 , wherein learning, via the machine learning algorithm, to identify the low resolution images of artifacts that are suitable for super-resolution generation by determining at least the context of each artifact, comprises:
identifying a location of each artifact on a respective specimen containing the artifact.
3. The method of claim 1 , wherein learning, via the machine learning algorithm, to identify the low resolution images of artifacts that are suitable for super-resolution generation by determining at least the context of each artifact, comprises:
determining a type of artifact undergoing inspection.
4. The method of claim 1 , wherein learning, via the machine learning algorithm, to identify the low resolution images of artifacts that are suitable for super-resolution generation by determining at least the context of each artifact, comprises:
mapping an input attribute vector comprising attributes of an artifact to a confidence that an input in the input attribute vector belongs to a suitability class.
5. The method of claim 4 , wherein the attributes comprise an artifact type, size, shape, composition, location on a specimen, or reference design.
6. The method of claim 4 , wherein the machine learning algorithm is a support vector machine.
7. The method of claim 1 , wherein the learning further comprises analyzing attributes of each artifact to classify each artifact.
8. The method of claim 1 , wherein a suitable artifact is a measure of a likelihood that the artifact can be used to product all or a portion of an accurate super-resolution image.
9. The method of claim 1 , further comprising:
identifying, by the computing system, an accuracy assessment of a super-resolution image of the target low resolution image based on one or more degrees of equivalence between the super-resolution image and the target low resolution image.
10. A super-resolution system comprising:
one or more processors; and
a memory having programming instructions stored thereon, which, when executed by the one or more processors, causes the super-resolution system to perform operations, comprising:
obtaining a training data set comprising a plurality of low resolution images of artifacts and an assigned suitability classification for each artifact;
generating a detection model for determining whether a detected artifact in a low resolution image is suitable for super-resolution generation by:
inputting the training data set into a machine learning algorithm; and
learning, via the machine learning algorithm, to identify low resolution images of artifacts that are suitable for super-resolution generation by determining at least a context of each artifact;
obtaining a target low resolution image of a specimen;
scanning the target low resolution image to identify a target artifact in the specimen;
analyzing, via the detection model, the target artifact to determine whether the target low resolution image is suitable for super-resolution imaging; and
upon determining, via the detection model, that the target low resolution image is suitable for super-resolution imaging, performing super-resolution imaging on the target low resolution image.
11. The super-resolution system of claim 10 , wherein learning, via the machine learning algorithm, to identify the low resolution images of artifacts that are suitable for super-resolution generation by determining at least the context of each artifact, comprises:
identifying a location of the artifact on a specimen containing the artifact.
12. The super-resolution system of claim 10 , wherein learning, via the machine learning algorithm, to identify the low resolution images of artifacts that are suitable for super-resolution generation by determining at least the context of each artifact, comprises:
determining a type of artifact undergoing inspection.
13. The super-resolution system of claim 10 , wherein learning, via the machine learning algorithm, to identify the low resolution images of artifacts that are suitable for super-resolution generation by determining at least the context of the artifact, comprises:
mapping an input attribute vector comprising attributes of an artifact to a confidence that an input in the input attribute vector belongs to a suitability class.
14. The super-resolution system of claim 13 , wherein the attributes comprise an artifact type, size, shape, composition, location on a specimen, or reference design.
15. The super-resolution system of claim 13 , wherein the machine learning algorithm is a support vector machine.
16. The super-resolution system of claim 10 , wherein the learning further comprises analyzing attributes of each artifact to classify the artifact.
17. The super-resolution system of claim 10 , wherein a suitable artifact is a measure of a likelihood that the artifact can be used to product all or a portion of an accurate super-resolution image.
18. The super-resolution system of claim 10 , wherein the operations further comprise:
identifying an accuracy assessment of a super-resolution image of the target low resolution image based on one or more degrees of equivalence between the super-resolution image and the target low resolution image.
19. A method comprising:
generating, by a computing system, a detection model for determining whether a detected artifact in a low resolution image is suitable for super-resolution generation by:
learning, via a machine learning algorithm, to identify low resolution images of artifacts that are suitable for super-resolution generation by determining at least a context of each artifact;
obtaining, by the computing system, a target low resolution image of a specimen;
scanning, by the computing system, the target low resolution image to identify at least one target artifact in the specimen;
analyzing, by the computing system via the detection model, the target artifact to determine whether the target low resolution image is suitable for super-resolution imaging; and
upon determining, by the computing via the detection model, that the target low resolution image is suitable for super-resolution imaging, performing super-resolution imaging on the target low resolution image.
20. The method of claim 19 , wherein generating, by the computing system, the detection model for determining whether the detected artifact in the low resolution image is suitable for the super-resolution generation, comprises:
generating a training data set comprising a plurality of low resolution images of artifacts and an assigned suitability classification for each artifact.