印刷电路板瑕疵检测与分类算法.pdfVIP

38 7 ( ) Vol. 38 No.7 2010 7 J. Huazhong Univ. of Sci. Tech. (Natural Science dition) Jul. 2010 陈晓辉 刘晓军 (华中科技大学 机械科学与工程学院, 湖北武汉430074) : ( PCB) ( AOI) .,, ., ., , PCB .: , , PCB. : ; ; ; ; ; : T P391. 4 : A :2010) A novel algorithm for inspecting and classifying the def ects of printed circuit boards Che n X i aoh ui L i u X i aoj un ( College of Mechanical Science and ngineering, Huazhong U niversity of Science and Technology, Wuhan 430074, China) Abstract: A novel automatic optical inspection( AOI) algorithm for detecting printed circuit board (PCB) defects based on the comparison of the character pattern of a reference and a test image is pres ented. Connected component analysis, which is basic method in morphology image processing tech niques, was used to extract the connected components,the character pattern, of the reference image. A concept of region of influence of connected component was mentioned to extract the character pat tern of the test image corresponding to that of the reference image. T he typical PCB defects such as open circuit, short, excessive short, missing conductor, pin hole, missing hole, breakout was detec ted and classified. A precise alignment for the reference and test image is not needed in this method, and it has a tolerance for some distortion of the image. T he algorithm can be applied in many kinds of PCB inspection. A result of inspected PCB defects is shown. Key words: automatic optical inspection; printed circuit board(PCB) ; reference image; test image; character pattern; region of influence of connected component