Image Processing Program, Image Processing Device, and Image Processing Method

Patent metadata

Field	Value
Publication number	US-2022318975-A1
Application number	US-201917596025-A
Country	US
Kind code	A1
Filing date	Jun 13, 2019
Priority date	Jun 13, 2019
Publication date	Oct 6, 2022
Grant date	—

Abstract

The purpose of the present invention is to provide a computer program for achieving die-to-database inspection at high speed and with few false reports, and a semiconductor inspection device using the same. To achieve this purpose, the present invention proposes: a computer program comprising an encoder layer that is configured to determine the features of a design data image, and a decoder layer that is configured to generate, on the basis of a variation in an image (inspection target image) obtained by photographing an inspection target pattern, a statistic pertaining to the brightness values of pixels from feature values output by the encoder layer, wherein die-to-database inspection with few false reports can be achieved by comparing the inspection target image and the statistic obtained from the decoder layer and pertaining to the brightness values, and thereby detecting a defect region in the image; and a semiconductor inspection device using the same.

First claim

1 .- 18 . (canceled)

19 . An image processing program for performing an inspection on a sample by using reference data of the sample, which is stored in a storage medium, and input data pertaining to the sample, the image processing program causing a processor to implement:
receiving the reference data; calculating a feature value pertaining to the sample on the basis of the reference data, by a convolutional neural network; and calculating a statistic indicating probability distribution of a value that is able to be taken by the input data on the basis of the feature value.

20 . The image processing program according to claim 19 , further causing a processor to implement:
determining necessity of learning for a parameter for calculating the feature value and a parameter for calculating the statistic; when the learning is determined to be needed, receiving the input data, comparing the statistic with the input data, and changing the parameter for calculating the feature value and the parameter for calculating the statistic in accordance with a result of the comparison; and when the learning is determined not to be needed, storing the parameter for calculating the feature value and the parameter for calculating the statistic as model data.

21 . The image processing program according to claim 19 , further causing a processor to implement:
receiving the input data; comparing the statistic with the input data; and evaluating the sample by using a result of the comparison.

22 . The image processing program according to claim 21 , wherein the evaluation of the sample is a defect inspection on the sample or a shape variation evaluation of the sample due to a process change.

23 . The image processing program according to claim 19 , wherein a value indicating the input data is a value indicating a shape or a physical property of the sample.

24 . The image processing program according to claim 21 , further causing a processor to implement:
displaying the statistic or a result of the evaluation.

25 . An image processing device for performing an inspection on a sample by using reference data of the sample, which is stored in a storage medium, and input data pertaining to the sample, the image processing device comprising:
a reference data receiving unit for receiving the reference data; a feature calculating unit for calculating a feature value pertaining to the sample on the basis of the reference data by a convolutional neural network; and a statistic calculating unit for calculating a statistic indicating probability distribution of a value that is able to be taken by the input data on the basis of the feature value.

26 . The image processing device according to claim 25 , further comprising:
a determining unit for determining necessity of learning for a parameter for calculating the feature value and a parameter for calculating the statistic; an input data receiving unit for receiving the input data; a comparing unit for comparing the statistic with the input data; and a changing and storing unit for changing and storing the parameter for calculating the feature value and the parameter for calculating the statistic, wherein, when the learning is determined to be needed, the comparing unit compares the statistic with the input data, and the changing unit changes the parameter for calculating the feature value and the parameter for calculating the statistic, and when the learning is determined not to be needed, the changing and storing unit stores the parameter for calculating the feature value and the parameter for calculating the statistic, as model data.

27 . The image processing device according to claim 26 , further comprising:
an evaluating unit for evaluating the sample by using a result of the comparison.

28 . The image processing device according to claim 27 , wherein the evaluation of the sample is a defect inspection on the sample or a shape variation evaluation of the sample due to a process change.

29 . The image processing device according to claim 25 , wherein a value indicating the input data is a value indicating a shape or a physical property of the sample.

30 . The image processing device according to claim 27 , further comprising:
a displaying unit for displaying the statistic or the result of the evaluation.

31 . An image processing method for performing an inspection on a sample by using reference data of the sample, which is stored in a storage medium, and input data pertaining to the sample, the image processing method comprising steps of:
receiving the reference data; calculating a feature value pertaining to the sample on the basis of the reference data by a convolutional neural network; and calculating a statistic indicating probability distribution of a value that is able to be taken by the input data on the basis of the feature value.

32 . The image processing method according to claim 31 , further comprising steps of:
determining necessity of learning for a parameter for calculating the feature value and a parameter for calculating the statistic; when the learning is determined to be needed, receiving the input data, comparing the statistic with the input data, and changing the parameter for calculating the feature value and the parameter for calculating the statistic in accordance with a result of the comparison; and when the learning is determined not to be needed, storing the parameter for calculating the feature value and the parameter for calculating the statistic as model data.

33 . The image processing method according to claim 31 , further comprising steps of:
receiving the input data; comparing the statistic with the input data; and evaluating the sample by using a result of the comparison.

34 . The image processing method according to claim 33 , wherein the evaluation of the sample is a defect inspection on the sample or a shape variation evaluation of the sample due to a process change.

35 . The image processing method according to claim 31 , wherein a value indicating the input data is a value indicating a shape or a physical property of the sample.

36 . The image processing method according to claim 33 , further comprising:
displaying the statistic or a result of the evaluation.

Assignees

• Hitachi High Tech Corp

Inventors

• Ouchi Masanori
• Shinoda Shinichi
• Toyoda Yasutaka
• Yumiba Ryou
• Shindo Hiroyuni

Classifications

• G06T2207/30148

  Semiconductor; IC; Wafer · CPC title

• G06T7/001 Primary

  using an image reference approach · CPC title

• G06T2207/20084

  Artificial neural networks [ANN] · CPC title

• G06T2207/20081

  Training; Learning · CPC title

• G06T2207/10061

  from scanning electron microscope · CPC title