Pattern alignment method based on consistency among local registration candidates for LSI wafer pattern inspection

Abstract

This paper reports an image-processing algorithm for robust inspection of LSI wafer patterns using SEM. In order to detect defects in a regular LSI pattern, a pair of long patterns are compared, blocked images are aligned, and defects are judged using the aligned images. The LSI wafer pattern is defined to consist of blank space, fine repetitive patterns, and unique patterns. Distortion of the SEM image is larger than the repetitive pattern pitch, requiring the system to keep track of the alignment in areas without pattern information or in blank space and mitigate the indeterminacy of repetitive patterns. To satisfy these requirements, a two-layer algorithm is proposed. The lower layer calculates registration candidates in each block, and the upper layer determines the correct registration route, i.e. the chain of the correct registration, using candidate information in all the related blocks. Experimental evaluations confirm that most pattern cases can be inspected correctly using the proposed SEM inspection system.

A new fuzzy based fast and high definition template matching algorithm

“…A lot of researching me search principle are aimed to incre reducing the quantity of the mat matching times in the image. Th projection transformation [6,7] and on contour area [8] es pasting production line r vision methods cannot w detection accurately due age with large sample piece a real-time on-line sample on the improved template is widely used in target h is simple and easy to precision [1,2], however its oved. The pros and cons of or to influence the speed of Appropriately reducing the the accuracy of similarity ove the speed of template ethods which modified the ase the searching speed by tching point and template he pyramid algorithm [3][4][5], matching algorithm based nt, and it is also an effective ng speed by improving the ation formula [9,10].…”