System and method for detecting defects on a wafer and related non-transitory computer-readable medium

A system configured to qualify a semiconductor wafer process, the system comprising:

• an inspection subsystem, configured to acquire scan data of a target region on a semiconductor wafer, wherein the target region comprises a plurality of conductive components on the semiconductor wafer and a plurality of defects in proximity to the conductive components or in the conductive components;
• and
• a computer system, configured to store a layout feature map of the conductive components of the semiconductor wafer, label the scan data with attributes from the layout feature map of conductive components, and determine a process healthiness from the labeled scan data by using a multi-dimension clustering method and a predetermined value,
• wherein the computer system is configured to group a plurality of entries of the scan data into entry groups through the multi-dimension clustering method,
• wherein the computer system is configured to expand each entry group based on grey levels of the scan data and determine entries out of a deviation range as the defects, and wherein the computer system is configured to apply different types of gains to different entry groups and filter different entry groups by different types of filters.

A system configured to qualify a semiconductor wafer process, the system comprising:

• an inspection subsystem, configured to acquire scan data of a target region on a semiconductor wafer, wherein the target region comprises a plurality of conductive components on the semiconductor wafer and a plurality of defects in proximity to the conductive components or in the conductive components;
• and
• a computer system, configured to store a layout feature map of the conductive components of the semiconductor wafer, label the scan data with attributes from the layout feature map of the conductive components, and determine the plurality of defects from the labeled scan data using a multi-dimension clustering method,
• wherein the computer system is configured to group a plurality of entries of the scan data into entry groups through the multi-dimension clustering method,
• wherein the computer system is configured to expand each entry group based on grey levels of the scan data and determine entries out of a deviation range as the defects, and wherein the computer system is configured to apply different types of gains to different entry groups and filter different entry groups by different types of filters.