Title :
Shape Feature Extraction for IC Wafer Circuit
Author :
Wu, Liming ; Wei, Qingchun ; Zhu, Jianlin ; Zailong, Qian
Author_Institution :
Fac. of Inf. Eng., Guangdong Univ. of Technol., Guangzhou, China
Abstract :
IC wafer image is composed of basic shape, mainly including lines, rectangles, rounds and ellipses. Firstly, this paper analyzes the shape of IC circuit image by the geometry of basic shape. When the images are moved and rotated, the moment has the same feature. Therefore, we can calculate the moment, the barycentric coordinate points and barycentric moment which are invariant in the same image. Secondly, this paper analyzes the theory of Hough transform and extracts the lines and rounds in the IC circuit image by modified Hough transform. Finally, the pads which are the rounds are detected, and the boundary which is the lines is detected. We can count the number of the lines and pads, and calculate the start and end of the lines, the center and radius of the pads. The experimental results demonstrate the modified Hough transform can effectively and accurately detect the rounds and lines in the IC wafer image.
Keywords :
Hough transforms; edge detection; electronic engineering computing; feature extraction; geometry; integrated circuits; IC wafer circuit; barycentric coordinate points; barycentric moment; boundary detection; geometry; line extraction; modified Hough transform; round detection; round extraction; shape feature extraction; Circuit noise; Computational geometry; Computer vision; Concurrent computing; Content based retrieval; Feature extraction; Image analysis; Parallel processing; Pattern recognition; Shape; Hough transform; IC wafer; line detection; round detection; shape feature;
Conference_Titel :
Information and Multimedia Technology, 2009. ICIMT '09. International Conference on
Conference_Location :
Jeju Island
Print_ISBN :
978-0-7695-3922-5
DOI :
10.1109/ICIMT.2009.24