摘要
针对X射线不均匀透视图像背景下的芯片焊接气泡检测问题,为了准确提取焊接气泡,特别是小气泡,提出了基于二维集合经验模式分解和基于灰度形态滤波的两种气泡检测算法.以大功率三极管芯片焊接气泡检测为案例,运用两种方法进行不均匀背景下气泡检测分析.结果表明,两种方法均能准确提取不均匀图像背景下的芯片气泡,而基于二维集合经验模式分解的芯片气泡检测方法更适合小气泡的检测,但要求图像噪声要小;基于灰度形态学的气泡检测方法更加全面且适于实时检测.
In order to realize the inspection of DIE welding voids in an uneven image background and etract small voids exactly,this paper presents two kinds of inspection methods which seperately use the 2-EEMD(two dimensional ensemble empirical mode decompositon) algorithm and the gray-scale morphology algorithm.A high power triode's DIE voids inspection is put forward for the qualification of the presented inspection methods.The results demostrate that the two inspection methods suceessfully extract the feature of the DIE welding voids in an uneven image background.The 2-EEMD welding voids inspection method is more suited to small voids inspection in nearly free-noise in X-ray image,and the gray-scale morphology inspection method is more suited in various situation and real-time applications.
出处
《焊接学报》
EI
CAS
CSCD
北大核心
2011年第10期65-68,116,共4页
Transactions of The China Welding Institution
基金
粤港关键领域重点突破招投标项目(东莞专项20092052-00013)
关键词
芯片气泡
集合经验模式分解
灰度形态学
DIE welding voids
ensemble empirical mode decomposition
gray-scale morphology
