公开(公告)号：US11821847B2

公开(公告)日：2023-11-21

申请号：US17380578

申请日：2021-07-20

Applicant: UNITED MICROELECTRONICS CORP.

Inventor： Cheng-Hsien Chen ,  Chia-Feng Hsiao ,  Chung-Hsuan Wu ,  Chen-Hui Huang ,  Nai-Ying Lo ,  En-Wei Tsui ,  Yung-Yu Yang ,  Chen-Hsuan Hung

IPC: G06T7/00 ,  G01N21/95 ,  G06T7/13

CPC classification number: G01N21/9505 ,  G06T7/0004 ,  G06T7/13 ,  G06T2207/30148

Abstract: A wafer backside defect detection method and a wafer backside defect detection apparatus are provided. The wafer backside defect detection method includes the following steps. A peripheral edge area of a wafer backside image that at least one notch is located is cropped off. Adjacent white pixels on the wafer backside image are connected to obtain a plurality of abnormal regions. If a total area of top N of the abnormal regions is more than 10% of an area of the wafer, it is deemed that the wafer has a roughness defect. N is a natural number. If the total area of the top N of the abnormal regions is less than 1% of the area of the wafer and a largest abnormal region of the abnormal regions is longer than a predetermined length, it is deemed that the wafer has a scratch defect.

公开(公告)号：US11644427B2

公开(公告)日：2023-05-09

申请号：US17100982

申请日：2020-11-23

Applicant: UNITED MICROELECTRONICS CORP.

Inventor： Chia-Feng Hsiao ,  Chung-Hsuan Wu ,  Shuo-Yu Chen ,  Nai-Ying Lo ,  Yi-Hui Tseng ,  Chen-Hui Huang ,  Yung-Yu Yang ,  Tzu-Ping Kao

IPC: G06K9/00 ,  G01N21/95 ,  H01L21/66 ,  G01N21/956 ,  G06T7/00 ,  G06T7/13 ,  G06V10/30 ,  G06V10/28 ,  G06V10/26 ,  G01N21/88 ,  G06T5/50 ,  G06V10/22

CPC classification number: G01N21/9503 ,  G01N21/8851 ,  G01N21/95607 ,  G06T5/50 ,  G06T7/001 ,  G06T7/0004 ,  G06T7/13 ,  G06V10/22 ,  G06V10/26 ,  G06V10/28 ,  G06V10/30 ,  H01L22/24 ,  G06T2207/20212 ,  G06T2207/20221 ,  G06T2207/30148

Abstract: An automatic detection method and an automatic detection system for detecting any crack on wafer edges are provided. The automatic detection method includes the following steps. Several wafer images of several wafers are obtained. The wafer images are integrated to create a templet image. Each of the wafer images is compared with the templet image to obtain a differential image. Each of the differential images is binarized. Each of the differential images which are binarized is de-noised. Whether each of the differential images has an edge crack is detected according to pattern of each of the differential images which are de-noised.