Decomposition-based classified ant colony optimization algorithm for scheduling semiconductor wafer fabrication system

A template-based vision system for the 100% inspection of wafer die surfaces has been developed. Design goals included a requirement for the detection of flaws as small as two thousandths of an inch on parts up to 8-in. wafer size. Each die is treated as one part of the whole wafer. One of the good dies is trained and kept as template die for the whole wafer. The die physical location data are generated, beforehand, from the wafer map supplied by the wafer manufacturers. A separate software package called the wafer map editor (WME) was developed to generate setup data needed during the runtime of the inspection process. The WME generates an updated wafermap with in-house defects after inspection of the wafer surface. Each unique die pattern in the wafer is defined as an object and these objects are grouped into user-defined categories, such as good die pads and defect die pads. The defect wafer map is used during the runtime inspection of die attach to avoid picking of bad dies.


  • Machine vision;
  • Wafer inspection;
  • Semiconductor;
  • Electronic wafermap;
  • Mean square error


For more information, please visit our, send us email at  or

Share this post