20 August 2015 Wafer defect inspection using component tree of SEM images
Author Affiliations +
This paper proposes a novel defect detection method using component tree of SEM images. The component tree contains rich information about the topological structure of images such as stiffness of intensity changes, area, and volume of lobes. The information is effective in detecting suspicious spot of the defects. A quasi-linear algorithm for constructing the component tree and computing those features is available. This paper modifies the original component tree algorithm to be suitable to the defect detection. One idea is excluding the pixels near the ground level in the initial stage of constructing the component tree. The other idea is to detect significant lobes by using both the volume and area attributes. The experiments performed with actual SEM images and printed electronics images showed promising results. For a 1000*1000 image, the proposed algorithm ran whole process in 1.36 seconds.
© (2015) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Sunghyon Kim, Sunghyon Kim, Minwoo Kim, Minwoo Kim, Il-suk Oh, Il-suk Oh, } "Wafer defect inspection using component tree of SEM images", Proc. SPIE 9556, Nanoengineering: Fabrication, Properties, Optics, and Devices XII, 95561J (20 August 2015); doi: 10.1117/12.2187411; https://doi.org/10.1117/12.2187411

Back to Top