Methods and systems to classify features in electronic designs
Methods for matching features in patterns for electronic designs include inputting a set of pattern data for semiconductor or flat panel displays, where the set of pattern data comprises a plurality of features. Each feature in the plurality of features is classified, where the classifying is based on a geometrical context defined by shapes in a region. The classifying uses machine learning techniques.
1. A method for matching features in patterns for electronic designs, the method comprising:
inputting a set of pattern data for semiconductor or flat panel displays, wherein the set of pattern data comprises a plurality of features;
classifying each feature in the plurality of features, wherein the classifying is based on a geometrical context defined by shapes in a region and wherein the classifying uses machine learning techniques;
creating a classification, the classification being a cluster of features from the classifying;
determining a mean cluster image from the classification; and
calculating a distance metric for each feature in the classification, wherein the distance metric indicates a deviation of each feature in the classification from the mean cluster image.
2. The method of claim 1 , further comprising compressing the input set of pattern data into compressed pattern data, wherein the classifying uses the compressed pattern data.
3. The method of claim 2 wherein the compressing uses an autoencoder.
4. The method of claim 3 wherein each encoded feature created by the autoencoder is an element in a vector.
5. The method of claim 1 wherein the classifying uses an autoencoder.
6. The method of claim 1 wherein the classifying allows a feature in the plurality of features to be in more than one classification.
7. The method of claim 1 wherein the set of pattern data comprises a set of questionable spots from mask inspection.
8. The method of claim 1 wherein the set of pattern data comprises simulated mask data enhanced by optical proximity correction (OPC).
9. The method of claim 1 wherein the set of pattern data comprises a set of reported errors from a geometric checker.
10. The method of claim 1 , wherein the distance metric is measured using a cosine distance of the features in the plurality of features within the classification.
11. The method of claim 1 , further comprising applying a Gaussian filter to a center of an image to give preference to features around the center.
12. The method of claim 1 , further comprising sorting the features in the classification by the distance metric, wherein the features in the classification with a largest distance have a higher priority for distinction.