DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 03/10/2026 has been entered.
Response to Arguments
Applicant’s arguments, see Remarks pages 9-17, filed 03/10/2026, with respect to the rejections of amended claims 1 and 17-18 under 35 U.S.C. 102(a)(1) have been fully considered and are moot in view of the new grounds of rejection (detailed in the rejections below) necessitated by Applicant's amendment to the claim(s).
Claim Interpretation
Note that according to the Federal Circuit’s 2004 Superguide v. DirecTV decision, “at least one of … and …” requires at least one instance of each and every item listed.
Claim(s) 1 and 17-18 recite(s): “wherein the processing parameter includes at least one element among circuit pattern shading, shape deformation, image resolution, and image noise of the sample.”
Claim(s) 1 and 17-18 recite(s): “at least one of the first learning image and the second learning image based on the design data.”
Claim(s) 5 recite(s): “the plurality of images of the plurality of image qualities are created by a change in at least one element among the circuit pattern shading, the shape deformation, the image resolution, and the image noise of the sample.”
Claim(s) 15 recite(s): “at a time of the learning, uses at least one of the first learning image and the second learning image as a tilt image obtained by observing a surface of the sample from diagonally above based on the design data;”
If Applicant intends for an interpretation of only one of these items being required for claim interpretation, Applicant can amend the claim language to, instead recite, “at least one of … or …”. In SuperGuide, the Federal Circuit held that the plain meaning of “at least one of A, B, and C” means: at least one of A, at least one of B, and at least one of C. The Court held that if the applicant intended “at least one of A, B, and C” to mean A, B, or C, they should have used “OR.” For the purposes of examination, the limitations are interpreted as disjunctive, thus requiring only one of the items, as disclosed by Pages 51: Lines 14-17, Page 43: Lines 10-12, and Page 57: Lines 10-15 of the Specification.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim(s) 1-2, 5, 12, 14, 16-18, and 20-23 is/are rejected under 35 U.S.C. 103 as being unpatentable over Harada et al. (US2018240225A1) hereinafter referenced as Harada, in view of Shinoda et al. (US2019228522A1) hereinafter referenced as Shinoda, Dou et al. (US2020074611A1) hereinafter referenced as Dou, and Zhang et al. (US2017148226A1) hereinafter referenced as Zhang.
Regarding claim 1, Harada discloses: A sample observation device comprising an imaging device and a processor (Harada: Figure 1), wherein the processor: stores design data on a sample in a storage resource (Harada: 0025: “an image storage unit that stores a degraded image having poor image quality and a high quality image having satisfactory image quality which are acquired at the same location of the sample”; Wherein images of a sample constitute design data);
selects a region of the sample (Harada: 0079: “ a region for acquiring a learning objective image from the sample wafer 1 is set (S 501 ). In this case, any given observation target region may be sampled and extracted from the sample wafer 1 , or may be randomly set within a sample surface.”);
creates a first learning image as a plurality of input images using a processing parameter (Harada: Figure 6; 0023: “aims to provide a sample observation device and a sample observation method which are capable of acquiring a high quality image with higher throughput by enabling the high quality image to be accurately estimated from degraded images including a degradation factor”), wherein the processing parameter includes at least one of element among circuit pattern shading, shape deformation, image resolution, and image noise of the sample (Harada: Figures 6 & 9; 0023: “and aims to provide a sample observation device and a sample observation method which are capable of acquiring a high quality image with higher throughput by enabling the high quality image to be accurately estimated from degraded images including a degradation factor (degraded resolution, a degraded SNR, blurring, and image shaking) occurring under a higher throughput imaging condition.”;
0107: “First, in a learning objective image pair acquisition step S 401 , a GUI 900 for setting the imaging condition of the degraded image and the high quality image is provided ( FIG. 9)...The items to be set in the column of the “imaging condition” 910 include the image resolution set in S 602 , the number of added frames set in S 601, a probe current of the electron beam 115 used by the SEM 101 irradiating the sample wafer 108 , an offset amount of the focus height of the electron beam 112 set in S 603 , and the scanning waiting time set in S 604.”; Wherein the Imaging Conditions in the GUI determine the degradation factors captured in the images.);
creates a second learning image as a target image using a parameter value designated by a user, and the parameter value represents a user preference relating to an ideal image quality (Harada: 0107: “First, in a learning objective image pair acquisition step S 401 , a GUI 900 for setting the imaging condition of the degraded image and the high quality image is provided ( FIG. 9)...The items to be set in the column of the “imaging condition” 910 include the image resolution set in S 602 , the number of added frames set in S 601, a probe current of the electron beam 115 used by the SEM 101 irradiating the sample wafer 108 , an offset amount of the focus height of the electron beam 112 set in S 603 , and the scanning waiting time set in S 604.”; Wherein the Imaging Conditions in the GUI determine the degradation factors captured in the high quality images, allowing for the user to select an ideal image quality based on the imaging condition parameter values.);
learns a model related to image quality conversion with the first learning image and the second learning image (Harada: Figure 4; 0025: “an arithmetic unit that calculates an estimation process parameter for estimating the high quality image from the degraded image by using the degraded image and the high quality image which are stored in the image storage unit,”; Wherein the arithmetic unit learns the estimation process parameter for image quality conversion);
acquires, as an observation image, a second captured image output by inputting a first captured image obtained by imaging the sample with the imaging device to the model in observing the sample (Harada: 0025: “a high quality image estimation unit that processes the degraded image obtained at a desired site of the sample which is obtained by causing the charged particle microscope to image the desired site of the sample, by using the estimation process parameter calculated by the arithmetic unit, and that estimates the high quality image obtained at the desired site, and an output unit that outputs the high quality image estimated by the high quality image estimation unit.”; Wherein the outputted high quality image is the observation image); and
creates at least one of the first learning image and the second learning image based on the design data (Claim limitation is interpreted according to SuperGuide claim interpretation recited above.) (Harada: 0079: “the SEM 101 images the region for acquiring the learning objective image, the degraded image is acquired (S 503 ), and the high quality image is acquired (S 504 ),”; Wherein the images of the sample, which are used to produce the degraded and high quality images, are design data).
Harada does not disclose expressly: wherein the processor: creates a first learning image of the region as a plurality of input images by separating the region into at least an upper layer pattern region and a lower layer pattern region, and creating a pattern-less region creating a pattern-less region using a processing parameter , wherein the processing parameter includes at least one element from among circuit pattern shading, shape deformation, image resolution, and image noise of the sample (Claim limitation is interpreted according to SuperGuide claim interpretation recited above.).
Shinoda discloses: wherein the processor: selects a region of a sample; creates a first learning image by separating the region into at least an upper layer pattern region and a lower layer pattern region (Shinoda: 0059: “The design data image generation unit 4 creates design data images shown in FIGS. 2B and 2C from the design data 2 corresponding to the SEM image 1 shown in FIG. 2A…In the design data image generation unit 4 , a region where there is a pattern is displayed with a white color, a region where there is no pattern is displayed with a black color, and a design data image is generated as a binary image.”;
0060: “Depending on a process, a plurality of layers of patterns may appear to be mixed in the SEM image. The design data has pattern information for each layer and the design data image is created by dividing the design data into upper and lower layers using the information.”); and creating a pattern-less region using a processing parameter, wherein the processing parameter includes at least one element from among circuit pattern shading, shape deformation, image resolution, or image noise of the sample (Shinoda: Figures 25A-25C; 0099-0100: “In the design data image generation units 4 and 10 , by changing a drawing method of the design drawing according to a photographing condition of an inspection image of photographing condition device information 32 or processing information of the device, an image close to the appearance of the inspection image can be created from the design data and prediction accuracy of the design data can be enhanced…in the case where the inspection target device is a pattern after etching processing as shown in FIG. 25A and the image of the inspection device is a BSE (reflected electron image) photographing image as shown in FIG. 25B, a brightness value of a groove (concave) portion of the pattern becomes smaller than a brightness value of the other (convex) portion. Therefore, at the time of drawing the design data, the brightness value of the groove portion of the pattern is set to be smaller than the brightness vale of the other portion, so that a design data image close to the appearance of the BSE photographing image can be generated.”; 0103: “Here, the photographing condition is a condition relating to photographing and shows a detection method (the BSE image, the SE image, and a combined image thereof), a frame accumulation number, a photographing magnification, or an image size, for example.”; Wherein the creation of design data images, which each comprise a pattern region and a no pattern region, constitute the creation of a pattern-less region using a processing parameter.)
Before the effective filing date of the claimed invention, it would have been obvious to a person of ordinary skill in the art to substitute the acquisition of degraded images based on imaging conditions disclosed by Harada with the imaging conditions based design data generation unit taught by Shinoda receiving imaging condition parameters though the GUI disclosed by Harada. The suggestion/motivation for doing so would have been “by changing a drawing method of the design drawing according to a photographing condition of an inspection image of photographing condition device information 32 or processing information of the device, an image close to the appearance of the inspection image can be created from the design data and prediction accuracy of the design data can be enhanced” (Shinoda: 0099). Further, one skilled in the art could have substituted the elements as described above by known methods with no change in their respective functions, and the substitution would have yielded nothing more than predictable results.
Harada in view of Shinoda does not disclose expressly: wherein the processor: creates a first learning image of the region as a plurality of input images by separating the region into at least an upper layer pattern region and a lower layer pattern region, adding the lower layer pattern region to the pattern-less region to create a multi-layer region, and adding the upper laver pattern region to the multi-layer region to create the first learning image of the region.
Dou discloses: wherein the processor: creates a first learning image of a region as a plurality of input images by separating the region into at least an upper layer pattern region and a lower layer pattern region, processing the upper and lower layer pattern regions, and merging the upper and lower layer pattern regions to create a multi-layer region, and thus the first learning image of the region (Dou: 0081: “In a case where the circuit pattern which is the object crosses a plurality of layers, teaching data may be created, in which the inside, the outside, and the contour line of each layer pattern are finely color-coded. In this case, a learning image based on the learning image coordinate list 104 is displayed on the GUI, and teaching data is superimposed on the image and created while visually being confirmed…In a true value assignment of a data set used for learning, the colors corresponding to attributions such as the number (indicating what layer it is) from an upper layer, the inside of the pattern, the outside of the pattern, the contour line, or the like are determined such that all data is unified...by using design data corresponding to the SEM image of the learning data, it is possible to obtain the number of attributions and the types of the attributions (the number of layers, the inside of the pattern, the outside of the pattern, and the contour line of the pattern) required for the true value assignment of the SEM image.”;
0090: “Here, it is conceivable that the generated model of the identifier is divided into a single layer, a multilayer, or the like. In this case, it is conceivable to select the model of the identifier using the design data. Moreover, when a learning data set is generated, the learning data set may be divided into a single layer, a multilayer, or the like using the design data so as to generate the learning data set.”).
Before the effective filing date of the claimed invention, it would have been obvious to a person of ordinary skill in the art to implement the known technique of color coding and layering pattern regions taught by Dou for the layering the single layer design images disclosed by Harada in view of Shinoda to create multilayer design images. The suggestion/motivation for doing so would have been “In a true value assignment of a data set used for learning, the colors corresponding to attributions such as the number (indicating what layer it is) from an upper layer, the inside of the pattern, the outside of the pattern, the contour line, or the like are determined such that all data is unified…in addition to the design data, it is conceivable that the type of attribution and the color corresponding to the number of attributions are determined by process information…it is conceivable to separately manage the color of the attribution of a structure such as a via-in-trench in which a via exists in a trench.” (Dou: 0081 & 0083; Wherein the images can be more granularly labeled). Further, one skilled in the art could have combined the elements as described above by known methods with no change in their respective functions, and the combination would have yielded nothing more than predictable results.
Harada in view of Shinoda and Dou does not disclose expressly: wherein the processor: trims the region based on the design data, creating a trimmed region; creates a first learning image of the trimmed region as a plurality of input images by separating the trimmed region into at least an upper layer pattern region and a lower layer pattern region, and adding the upper laver pattern region to the multi-layer region to create the first learning image of the trimmed region; and creates a second learning image of the trimmed region as a target image using a parameter value designated by a user.
Thus, Harada in view of Shinoda and Dou does not disclose expressly: the trimming of a region based on the design data, such that the first learning image and the second learning image are processed to pertain to the same trimmed region.
Zhang discloses: the generation of a training dataset comprised of design data and real images, wherein the generation of the training dataset consists of the alignment and image cropping of the CAD and real images (Zhang: 0072: “the computer subsystem(s) are configured for generating a training dataset used for training the generative model, and the training dataset includes a set of pairs of a portion of other design information and an actual image generated for the portion of the other design information. The other design information may include any of the design information described herein. The computer subsystem(s) may, therefore, prepare a training dataset. In this step, a training dataset may be generated by alignment and image cropping of CAD and real images, resulting in a collection of pairs of aligned CAD and real image (e.g., SEM or optical image), wherein “real” images are those that are generated by imaging a physical specimen on which the design information has been formed.”).
Before the effective filing date of the claimed invention, it would have been obvious to a person of ordinary skill in the art to implement the algorithms for aligning and cropping the CAD and real images taught by Zhang for the generation of the training dataset disclosed by Harada in view of Shinoda and Dou. The suggestion/motivation for doing so would have been “The CAD may be rendered as a binary image and uneven pixel size may be restored before alignment. The pixel size may be “uneven” due to image distortion from hardware. For example, in an electron beam image, due to the instability of the electron beam, a pixel at a first location on a specimen could represent a 10 nm by 10 nm area on the specimen while another pixel at a second location on the specimen (which may be relatively close to the first location) may represent a 11 nm by 11 nm area on the specimen, while the expected pixel size is 10.5 nm by 10.5 nm” (Zhang: 0073; Wherein the alignment and cropping operations serve to correct distortions between the images due to hardware prior to training.). Further, one skilled in the art could have combined the elements as described above by known methods with no change in their respective functions, and the combination would have yielded nothing more than predictable results. Therefore, it would have been obvious to combine Harada in view of Shinoda and Dou with Zhang to obtain the invention as specified in claim 1.
Regarding claim 2, Harada in view of Shinoda, Dou, and Zhang discloses: The sample observation device according to claim 1, wherein the processor: creates the first learning image based on the design data (Shinoda: 0059-0060: “The design data image generation unit 4 creates design data images shown in FIGS. 2B and 2C from the design data 2 corresponding to the SEM image 1 shown in FIG. 2A…Depending on a process, a plurality of layers of patterns may appear to be mixed in the SEM image. The design data has pattern information for each layer and the design data image is created by dividing the design data into upper and lower layers using the information.”; Wherein the design data images are generated based on the imaging conditions disclosed by Harada.); and creates the second learning image based on the design data (Harada: 0088: “The process flow of acquiring the high quality image (S 504) is also the same as the process flow of acquiring the degraded image described in FIG. 6. However, in order to acquire the high quality image, the imaging condition is set so that a sufficiently less degraded image can be obtained.”; Wherein the images of the wafer are design data).
Regarding claim 5, Harada in view of Shinoda, Dou, and Zhang discloses: The sample observation device according to claim 1, wherein the first learning image includes a plurality of images of a plurality of image qualities (Harada: 0023: “and aims to provide a sample observation device and a sample observation method which are capable of acquiring a high quality image with higher throughput by enabling the high quality image to be accurately estimated from degraded images including a degradation factor (degraded resolution, a degraded SNR, blurring, and image shaking) occurring under a higher throughput imaging condition.”; 0055: “the degraded image corresponding to each degradation factor is acquired, and the degraded image is used for learning.”), and the plurality of images of the plurality of image qualities are created by a change in at least one element among the circuit pattern shading, the shape deformation, the image resolution, and the image noise of the sample (Claim limitation is interpreted according to SuperGuide claim interpretation recited above.) (Harada: 0053: “Examples of the degraded image include an image having low resolution (fewer pixels), an image having a low a signal to noise ratio (SNR), a blurred image generated due to focus misalignment, a shaken image generated due to a shaken stage which holds a sample when imaged.”)
(Shinoda: Figures 25A-25C; 0099: “In the design data image generation units 4 and 10 , by changing a drawing method of the design drawing according to a photographing condition of an inspection image of photographing condition device information 32 or processing information of the device, an image close to the appearance of the inspection image can be created from the design data and prediction accuracy of the design data can be enhanced.”).
Regarding claim 12, Harada in view of Shinoda, Dou, and Zhang discloses: The sample observation device according to claim 1, wherein the processor measures a circuit pattern dimension of the sample using the observation image in observing the sample (Harada: 0118: “The high quality image estimated through the above-described technique may be used so as to perform defect detection…and circuit pattern measurement of detected defects.”).
Regarding claim 14, Harada in view of Shinoda, Dou, and Zhang discloses: The sample observation device according to claim 1, wherein the processor specifies a position of a defect of the sample using the observation image by the second captured image output (Harada: 0118: “The high quality image estimated through the above-described technique may be used so as to perform defect detection, or observation, classification, and circuit pattern measurement of detected defects.”; 0121: “the present embodiment intends to realize improved throughput in a case of sequentially observing a plurality of defects.”) by inputting the first captured image obtained by imaging defect coordinates indicated by defect position information to the model in observing the sample (Harada: 0124: “the stage 109 is controlled and moved by the stage control unit 201 so that the coordinate of the position of the learning target defect read in order to image the learning target defect in S 1202 is included in the imaging visual field of the SEM 101 (S 1203: corresponding to S 502 in FIG. 5).”).
Regarding claim 16, Harada in view of Shinoda, Dou, and Zhang discloses: The sample observation device according to claim 1, wherein the processor causes the first learning image or the second learning image created based on the design data to be displayed on a screen (Harada: Figure 11B; 0112: “a degraded image 1121 corresponding to the designated image ID and an estimation result 1122 obtained from the degraded image 1121 through the estimation process, and a high quality image 1123 are displayed.”).
As per claim(s) 17, arguments made in rejecting claim(s) 1 are analogous.
As per claim(s) 18, arguments made in rejecting claim(s) 1 are analogous. Additionally, Figures 1 and 2 of Harada disclose a sample observation device containing a control system unit with a control unit and storage unit implying a “computer system in a sample observation device including an imaging device”.
Regarding claim 20, Harada in view of Shinoda, Dou, and Zhang discloses: The sample observation device according to claim 1, wherein the design data includes an image of the sample (Shinoda: 0042-0043: “a simulator 2404 that simulates a result of the pattern on the basis of the design data of the semiconductor device and a manufacturing condition of a semiconductor manufacturing device, and a storage medium 2405 that stores the design data in which layout data of the semiconductor device or the manufacturing condition is registered are connected to the network. The design data is expressed in, for example, a GDS format or an OASIS format and is stored in a predetermined format. The design data may be of any type as long as software that displays the design data can display the format and can be handled as graphic data.”), and wherein the processor: creates the first learning image based on the image of the design data (Shinoda: Figures 25A-25C; 0099: “In the design data image generation units 4 and 10 , by changing a drawing method of the design drawing according to a photographing condition of an inspection image of photographing condition device information 32 or processing information of the device, an image close to the appearance of the inspection image can be created from the design data and prediction accuracy of the design data can be enhanced.”).
Regarding claim 21, Harada in view of Shinoda, Dou, and Zhang discloses: The sample observation device according to claim 1, wherein the design data includes an image of the sample (Shinoda: 0042-0043: “a simulator 2404 that simulates a result of the pattern on the basis of the design data of the semiconductor device and a manufacturing condition of a semiconductor manufacturing device, and a storage medium 2405 that stores the design data in which layout data of the semiconductor device or the manufacturing condition is registered are connected to the network. The design data is expressed in, for example, a GDS format or an OASIS format and is stored in a predetermined format. The design data may be of any type as long as software that displays the design data can display the format and can be handled as graphic data.”), and wherein the processor: creates the second learning image based on the image in the design data (Shinoda: 0060: “a plurality of layers of patterns may appear to be mixed in the SEM image. The design data has pattern information for each layer and the design data image is created by dividing the design data into upper and lower layers using the information.”)
(Harada: 0088: “The process flow of acquiring the high quality image (S 504) is also the same as the process flow of acquiring the degraded image described in FIG. 6. However, in order to acquire the high quality image, the imaging condition is set so that a sufficiently less degraded image can be obtained.”; Wherein the high-quality images and design data information share layout and structural information regarding the processed wafer).
Regarding claim 22, Harada in view of Shinoda, Dou, and Zhang discloses: The sample observation device according to claim 20, wherein the processor trims a region of an image of a corresponding position from the image in the design data (Zhang: 0072: “a training dataset may be generated by alignment and image cropping of CAD and real images, resulting in a collection of pairs of aligned CAD and real image (e.g., SEM or optical image), wherein “real” images are those that are generated by imaging a physical specimen on which the design information has been formed.”).
Regarding claim 23, Harada in view of Shinoda, Dou, and Zhang discloses: The sample observation device according to claim 21, wherein the processor trims a region of an image of a corresponding position from the image in the design data (Zhang: 0072: “a training dataset may be generated by alignment and image cropping of CAD and real images, resulting in a collection of pairs of aligned CAD and real image (e.g., SEM or optical image), wherein “real” images are those that are generated by imaging a physical specimen on which the design information has been formed.”).
Claim(s) 10 is rejected under 35 U.S.C. 103 as being unpatentable over Harada in view of Shinoda, Dou, and Zhang, and further in view of Gupta et al. (US 20180293721 A1) hereinafter referenced as Gupta.
Regarding claim 10, Harada in view of Shinoda, Dou, and Zhang discloses: The sample observation device according to claim 1.
Harada in view of Shinoda, Dou, and Zhang does not disclose expressly: wherein, in creating the second learning image based on the design data, the processor creates an edge image in which a pattern contour line of the sample is drawn from a region of the design data.
Gupta discloses creating an edge image in which pattern contour lines of the sample are drawn from images of the sample taken (Gupta: Figure 7; 0015: “The second learning based model is configured for generating actual contours for the patterns in at least one of the acquired images of the patterns formed on the specimen input to the second learning based model by the one or more computer subsystems.”).
Before the effective filing date of the claimed invention, it would have been obvious to a person of ordinary skill in the art to process the high-quality images disclosed by Harada in view of Shinoda, Dou, and Zhang through the second learning model configured to generate the contours of the pattern present in the image taught by Gupta. The suggestion/motivation for doing so would have been “Currently used methods for CD-SEMs have several challenges such as that they are slow, they require careful setup for each site and a knowledge of which sites to measure, and their results need to be interpreted further downline.” (Gupta: 0008). Further, one skilled in the art could have combined the elements as described above by known methods with no change in their respective functions, and the combination would have yielded nothing more than predictable results. Therefore, it would have been obvious to combine Harada in view of Shinoda, Dou, and Zhang with Gupta to obtain the invention as specified in claim 10.
Claim(s) 11 is rejected under 35 U.S.C. 103 as being unpatentable over Harada in view of Shinoda, Dou, Zhang, and Gupta, and further in view of Tsankashvili (Tsankashvili; N, “Comparing Edge Detection Methods”).
Regarding claim 11, Harada in view of Shinoda, Dou, Zhang, and Gupta discloses: The sample observation device according to claim 10.
Harada in view of Shinoda, Dou, Zhang, and Gupta does not disclose expressly: wherein the processor: in creating the edge image, creates a plurality of edge images in which direction-specific pattern contour lines in a plurality of directions are drawn from a region of the design data; and at a time of the learning, learns the model with the first learning image and a plurality of images corresponding to the plurality of edge images as the second learning image.
Tsankashvili discloses in creating the edge image, creates a plurality of edge images in which direction-specific pattern contour lines in a plurality of directions are drawn from a region (Tsankashvili: Page 2: “The Sobel is one of the most commonly used edge detectors. It is based on convolving the image with a small, separable, and integer valued filter in horizontal and vertical direction and is therefore relatively inexpensive in terms of computations.”; Wherein the horizontal and vertical contour images are created individually through the Sobel operator computations), of which the plurality of edge images are then combined into a single image (Tsankashvili: Page 4:
PNG
media_image1.png
86
545
media_image1.png
Greyscale
; Wherein the vertical and horizontal image gradients are combined into a single image).
Before the effective filing date of the claimed invention, it would have been obvious to a person of ordinary skill in the art to substitute the algorithms present in the model configured to generate the contours present in the image disclosed by Harada in view of Shinoda, Dou, Zhang, and Gupta with the Sobel Operator algorithm configured to generate a plurality of edge images taught by Tsankashvili. The suggestion/motivation for doing so would have been “The Sobel edge enhancement filter has the advantage of providing differentiating (which gives the edge response) and smoothing (which reduces noise) concurrently” (Tsankashvili: Page 2). Further, one skilled in the art could have substituted the elements as described above by known methods with no change in their respective functions, and the substitution would have yielded nothing more than predictable results. Therefore, it would have been obvious to combine Harada in view of Shinoda, Dou, Zhang, and Gupta with Tsankashvili to obtain the invention as specified in claim 11.
Claim(s) 13 is rejected under 35 U.S.C. 103 as being unpatentable over Harada in view of Shinoda, Dou, and Zhang, and further in view of Grama et al. (US20210343001A1) hereinafter referenced as Grama.
Regarding claim 13, Harada in view of Shinoda, Dou, and Zhang discloses: The sample observation device according to claim 1.
Harada in view of Shinoda, Dou, and Zhang does not disclose expressly: wherein the processor specifies an imaging position of the first captured image by performing alignment between the observation image and the design data using the observation image in observing the sample.
Grama discloses: wherein the processor specifies an imaging position of the first captured image by performing alignment between the observation image and the design data using the observation image in observing the sample (Grama: 0116: “the higher resolution images may be aligned to the design for the specimen with greater accuracy and precision by an align-to-design method or algorithm such as pixel-to-design alignment (PDA) algorithms used by some tools commercially available from KLA…the information determined for the specimen using the higher resolution images may include coordinates of defects in design data space…”; Wherein the high quality image is aligned to the design and may display defects based on the coordinates of the design data).
Before the effective filing date of the claimed invention, it would have been obvious to a person of ordinary skill in the art to align the observation image disclosed in Harada in view of Shinoda, Dou, and Zhang based on design data through a align-to-design method as taught by Grama. The suggestion/motivation for doing so would have been “The information about the position of the defects relative to the design…may also be used to determine additional information for the specimen such as information about weak or hot points in the design (i.e., locations where defects occur or tend to occur repeatedly in multiple instances of the same portion of the design).” (Grama: 0116; Wherein the design data for the wafer may be stored with information for spots where defects may frequently occur). Further, one skilled in the art could have combined the elements as described above by known methods with no change in their respective functions, and the combination would have yielded nothing more than predictable results. Therefore, it would have been obvious to combine Harada in view of Shinoda, Dou, and Zhang with Grama to obtain the invention as specified in claim 13.
Claim(s) 15 is rejected under 35 U.S.C. 103 as being unpatentable over Harada in view of Shinoda, Dou, and Zhang, and further in view of Bhattacharyya et al. (US2021133989A1) hereinafter referenced as Bhattacharyya.
Regarding claim 15, Harada in view of Shinoda, Dou, and Zhang discloses: The sample observation device according to claim 1.
Harada in view of Shinoda, Dou, and Zhang does not disclose expressly: wherein the processor: at a time of the learning, uses at least one of the first learning images and second learning image as a tilt image obtained by observing a surface of the sample from diagonally above based on the design data (Claim limitation is interpreted according to SuperGuide claim interpretation recited above.); and in observing the sample, acquires, as the observation image, a tilt image as the second captured image output by inputting a tilt image obtained by imaging the surface of the sample from diagonally above with the imaging device to the model as the first captured image.
Bhattacharyya discloses a wafer inspection tool (Bhattacharyya: Figure 6) configured to generate images of a wafer at oblique angles (Bhattacharyya: 0062: “Although the electron column 601 is shown in FIG. 6 as being configured such that the electrons are directed to the wafer 604 at an oblique angle of incidence and are scattered from the wafer 604 at another oblique angle, the electron beam may be directed to and scattered from the wafer 604 at any suitable angles.”) for the training of a model (Bhattacharyya: Figure 4; 0040: “A generator network G can generate a “fake” image G(x) from the input (e.g., a SEM image)…G can be trained to minimize error between a real and fake image and to fool the discriminator, which can be represented by minimizing |y−G(x)|+log(1−D(G(x))).”) and observation of a sample (Bhattacharyya: 0012: “The system comprises an SEM including an electron emitter, a detector, and a stage for holding a sample. The SEM is configured to obtain an SEM image of the sample…The defects of the thresholded probability map and the signal-to-noise-ratio defects of the thresholded defect map are filtered using a broadband-plasma-based property to produce defect-of-interest clusters.”; Wherein the defect-of interest clusters are produced from observing the SEM image of the sample).
Before the effective filing date of the claimed invention, it would have been obvious to a person of ordinary skill in the art to substitute the imaging of the sample with the immovable imaging microscope disclosed by Harada in view of Shinoda, Dou, and Zhang with imaging the sample with the directable electron microscope taught by Bhattacharyya in order to capture the high-quality learning images and observation images disclosed in Harada in view of Shinoda, Dou, and Zhang as tilt images. The suggestion/motivation for doing so would have been to increase the possibilities of capturing and detecting the defects present in a sample through the imaging at various angles. Further, one skilled in the art could have combined the elements as described above by known methods with no change in their respective functions, and the combination would have yielded nothing more than predictable results. Therefore, it would have been obvious to combine Harada in view of Shinoda, Dou, and Zhang with Bhattacharyya to obtain the invention as specified in claim 15.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ANTHONY J RODRIGUEZ whose telephone number is (703)756-5821. The examiner can normally be reached Monday-Friday 10am-7pm.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Sumati Lefkowitz can be reached at (571) 272-3638. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/ANTHONY J RODRIGUEZ/Examiner, Art Unit 2672
/SUMATI LEFKOWITZ/Supervisory Patent Examiner, Art Unit 2672
Prosecution Insights