System, method and apparatus for macroscopic inspection of reflective specimens
An inspection apparatus includes a specimen stage, one or more imaging devices and a set of lights, all controllable by a control system. By translating or rotating the one or more imaging devices or specimen stage, the inspection apparatus can capture a first image of the specimen that includes a first imaging artifact to a first side of a reference point and then capture a second image of the specimen that includes a second imaging artifact to a second side of the reference point. The first and second imaging artifacts can be cropped from the first image and the second image respectively, and the first image and the second image can be digitally stitched together to generate a composite image of the specimen that lacks the first and second imaging artifacts.
1. A method, comprising:
generating, by a computing system, a training data set for training a machine learning model to generate an illumination profile for illuminating a specimen under examination in a macro inspection system, the training data set comprising images of known specimens and illumination profile data corresponding to the images of known specimens;
training, by the computing system, the machine learning model to generate illumination profiles for illuminating specimens under examination based on the training data set, wherein the machine learning model learns features of the known specimens and correlates the learned features with the illumination profile data;
determining, by the computing system, that the machine learning model has achieved a threshold level of accuracy in generating the illumination profiles for illuminating specimens under examination; and
based on the determining, deploying, by the computing system, the machine learning model in a macro inspection environment.
2. The method of claim 1 , wherein the training data set further comprises:
non-image data identifying known specimens and features of the known specimens.
3. The method of claim 1 , wherein the illumination profile data in the training data set comprises one or more of an activation value, an intensity value, or a color value of each light utilized for examination.
4. The method of claim 1 , wherein the training data set further comprises:
a region of interest for each of the known specimens.
5. The method of claim 1 , wherein the training data set further comprises:
an indication of a particular stage of manufacturing for the known specimen undergoing the examination.
6. The method of claim 1 , wherein training, by the computing system, the machine learning model to generate the illumination profiles comprises:
training the machine learning model to predict one or more of an activation value, an intensity value, or a color value of each light utilized for examination.
7. The method of claim 1 , wherein the training data set comprises:
a distance of the known specimen to a lens of the macro inspection system.
8. A non-transitory computer readable medium comprising one or more sequences of instructions, which, when executed by one or more processors, causes a computing system to perform operations comprising:
generating, by the computing system, a training data set for training a machine learning model to generate an illumination profile for illuminating a specimen under examination in a macro inspection system, the training data set comprising images of known specimens and illumination profile data corresponding to the images of known specimens;
training, by the computing system, the machine learning model to generate illumination profiles for illuminating specimens under examination based on the training data set, wherein the machine learning model learns features of the known specimens and correlates the learned features with the illumination profile data;
determining, by the computing system, that the machine learning model has achieved a threshold level of accuracy in generating the illumination profiles for illuminating specimens under examination; and
based on the determining, deploying, by the computing system, the machine learning model in a macro inspection environment.
9. The non-transitory computer readable medium of claim 8 , wherein the training data set further comprises:
non-image data identifying known specimens and features of the known specimens.
10. The non-transitory computer readable medium of claim 8 , wherein the illumination profile data in the training data set comprises one or more of an activation value, an intensity value, or a color value of each light utilized for examination.
11. The non-transitory computer readable medium of claim 8 , wherein the training data set further comprises:
a region of interest for each of the known specimens.
12. The non-transitory computer readable medium of claim 8 , wherein the training data set further comprises:
an indication of a particular stage of manufacturing for the known specimen undergoing the examination.
13. The non-transitory computer readable medium of claim 8 , wherein training, by the computing system, the machine learning model to generate the illumination profiles comprises:
training the machine learning model to predict one or more of an activation value, an intensity value, or a color value of each light utilized for examination.
14. The non-transitory computer readable medium of claim 8 , wherein the training data set comprises:
a distance of the known specimen to a lens of the macro inspection system.
15. A system comprising:
a processor; and
a memory having programming instructions stored thereon, which, when executed by the processor, causes a computing system to perform operations comprising:
generating, by the computing system, a training data set for training a machine learning model to generate an illumination profile for illuminating a specimen under examination in a macro inspection system, the training data set comprising images of known specimens and illumination profile data corresponding to the images of known specimens;
training, by the computing system, the machine learning model to generate illumination profiles for illuminating specimens under examination based on the training data set, wherein the machine learning model learns features of the known specimens and correlates the learned features with the illumination profile data;
determining, by the computing system, that the machine learning model has achieved a threshold level of accuracy in generating the illumination profiles for illuminating specimens under examination; and
based on the determining, deploying, by the computing system, the machine learning model in a macro inspection environment.
16. The system of claim 15 , wherein the training data set further comprises:
non-image data identifying known specimens and features of the known specimens.
17. The system of claim 15 , wherein the illumination profile data in the training data set comprises one or more of an activation value, an intensity value, or a color value of each light utilized for examination.
18. The system of claim 15 , wherein the training data set further comprises:
a region of interest for each of the known specimens.
19. The system of claim 15 , wherein the training data set further comprises:
an indication of a particular stage of manufacturing for the known specimen undergoing the examination.
20. The system of claim 15 , wherein training, by the computing system, the machine learning model to generate the illumination profiles comprises:
training the machine learning model to predict one or more of an activation value, an intensity value, or a color value of each light utilized for examination.