-
公开(公告)号:US20240320805A1
公开(公告)日:2024-09-26
申请号:US18612094
申请日:2024-03-21
Applicant: SAMSUNG ELECTRONICS CO., LTD.
Inventor: Nohong Kwak , Donyun Kim , Jiwon Kang , Kihyun Kim
CPC classification number: G06T5/70 , G06T7/0004 , G06V10/44 , G06T2207/20081 , G06T2207/30148
Abstract: There is provided an image denoising method including extracting a noise patch from a noisy image, outputting a noise parameter by inputting the noise patch to a noise parameter estimation network (NPE-net), generating imitated virtual noise based on the output noise parameter, generating a noisier image by adding the imitated virtual noise to the noisy image, training a denoise deep learning model by inputting the noisy image and the noisier image as a pair to the denoise deep learning model, inputting the noisy image to the trained denoise deep learning model, and outputting a denoise image obtained by removing noise from the noisy image by the trained denoise deep learning model.
-
公开(公告)号:US12205267B2
公开(公告)日:2025-01-21
申请号:US17737494
申请日:2022-05-05
Inventor: Nohong Kwak , Donyun Kim , Kihyun Kim , Yunhyoung Nam
IPC: G06V10/762 , G06T7/00
Abstract: A pattern inspection method includes converting sample patterns of a sample image into training images, extracting feature values of the training patterns, setting feature vectors of the training patterns on the basis of the feature values, converting the feature vectors into Gaussian vectors, clustering the Gaussian vectors, thereby sorting the Gaussian vectors into clusters, selecting a select vector from each of the clusters, storing, as a pattern model, the training pattern corresponding to the select vector in a database, converting a target pattern of a target image into an inspection pattern on the basis of the pattern model, and inspecting the inspection pattern.
-
公开(公告)号:US11740073B2
公开(公告)日:2023-08-29
申请号:US17463499
申请日:2021-08-31
Inventor: Jooho Kim , Donyun Kim , Yunhyoung Nam , Seungjin Lee , Dawoon Choi
CPC classification number: G01B11/24 , G06T7/60 , G06V10/267 , H01J37/28 , H01J2237/24578 , H01J2237/2803
Abstract: A method of measuring a critical dimension (CD) includes forming a plurality of patterns in a substrate, creating first to n-th images, where n is a natural number greater than 1, for first to n-th areas in the substrate, respectively, where the first to n-th areas do not overlap with each other, where each of the first to n-th areas comprising at least some of the plurality of patterns, creating a merged image for the first to n-th images, and measuring a CD for a measurement object from the plurality of patterns using the merged image. The merged image has a higher resolution than each of the first to n-th images.
-
-