芯片焊盘自动识别与定位方法的研究  

Research on the Way of Automatic Recognizing and Positioning Die Pads

作　　者：吴川[1] 朱明[1] 杨冬[2] 

机构地区：[1]中国科学院长春光学精密机械与物理研究所图像处理室,长春130033 [2]长春工业大学计算机科学与工程学院电子教研室,长春130012

出　　处：《计算机辅助设计与图形学学报》2005年第8期1673-1677,共5页Journal of Computer-Aided Design & Computer Graphics

摘　　要：为了准确找到芯片焊盘的位置,首先对已有隶属函数进行改造,用一种新的基于最大模糊熵原则的阈值分割法对图像进行二值处理;然后采用计算图像重心的方法得到芯片的大致位置;最后针对芯片上各焊盘的位置特点,以焊盘为模板,通过计算模板图像和目标图像之间的汉明距离来识别各焊盘的位置.实验结果表明:文中方法能准确地识别出芯片焊盘位置.In order to detect the locations of the die pads, membership function is firstly reconstructed.Then a new threshold segmentation method based on maximum fuzzy entropy is proposed to obtain a binary image. Third, the rough position is fitted by resembling the barycenter of the image. Finally, according to the characteristic of positions of the die pads, their locations are fixed on through calculating the Hamming distance between the template image and the subject image. Experimental results demonstrate that the proposed approach can detect the locations of the die pads exactly.

关 键 词：隶属函数 最大模糊熵 芯片 汉明距离 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]