-
公开(公告)号：US09355440B1
公开(公告)日：2016-05-31
申请号：US13649080
申请日：2012-10-10
Applicant: KLA-Tencor Corporation
Inventor： Haiguang Chen , Michael D. Kirk , Stephen Biellak , Jaydeep Sinha
CPC classification number: G06T7/0012 , G01N21/8851 , G01N21/95623 , G01N2021/8887 , G06K9/4633 , G06K2209/19 , G06T7/0004 , G06T7/60 , G06T2207/20056 , G06T2207/20061 , G06T2207/30148
Abstract: Methods and systems for detection of selected defects in relatively noisy inspection data are provided. One method includes applying a spatial filter algorithm to inspection data acquired across an area on a substrate to determine a first portion of the inspection data that has a higher probability of being a selected type of defect than a second portion of the inspection data. The selected type of defect includes a non-point defect. The inspection data is generated by combining two or more raw inspection data corresponding to substantially the same locations on the substrate. The method also includes generating a two-dimensional map illustrating the first portion of the inspection data. The method further includes searching the two-dimensional map for an event that has spatial characteristics that approximately match spatial characteristics of the selected type of defect and determining if the event corresponds to a defect having the selected type.
Abstract translation: 提供了用于检测相对噪声检查数据中选定缺陷的方法和系统。 一种方法包括将空间滤波器算法应用于跨越衬底上的区域获取的数据，以确定检查数据的第一部分与检查数据的第二部分相比具有较高选择类型的缺陷概率。 所选择的缺陷类型包括非点缺陷。 通过组合对应于基板上基本上相同的位置的两个或更多个原始检查数据来生成检查数据。 该方法还包括生成示出检查数据的第一部分的二维映射。 该方法还包括搜索具有近似匹配所选类型的缺陷的空间特征的空间特征的事件的二维地图，并且确定该事件是否对应于具有所选类型的缺陷。
-
公开(公告)号：US09646379B1
公开(公告)日：2017-05-09
申请号：US15139315
申请日：2016-04-26
Applicant: KLA-Tencor Corporation
Inventor： Haiguang Chen , Michael D. Kirk , Stephen Biellak , Jaydeep Sinha
CPC classification number: G06T7/0012 , G01N21/8851 , G01N21/95623 , G01N2021/8887 , G06K9/4633 , G06K2209/19 , G06T7/0004 , G06T7/60 , G06T2207/20056 , G06T2207/20061 , G06T2207/30148
Abstract: Methods and systems for detection of selected defects in relatively noisy inspection data are provided. One method includes applying a spatial filter algorithm to inspection data acquired across an area on a substrate to determine a first portion of the inspection data that has a higher probability of being a selected type of defect than a second portion of the inspection data. The selected type of defect includes a non-point defect. The inspection data is generated by combining two or more raw inspection data corresponding to substantially the same locations on the substrate. The method also includes generating a two-dimensional map illustrating the first portion of the inspection data. The method further includes searching the two-dimensional map for an event that has spatial characteristics that approximately match spatial characteristics of the selected type of defect and determining if the event corresponds to a defect having the selected type.
-
Search Results
2
records
(took 0.024 seconds)
