气囊封条贴片缺陷检测系统设计  

Design of patch defect detection system for airbag seals

作　　者：张相胜 王甫姜 邬鹏 潘丰 ZHANG Xiang-sheng;WANG Fu-jiang;WU Peng;PAN Feng(Key Laboratory of Advanced Process Control for Light Industry,Ministry of Education, Jiangnan University,Wuxi 214122,China)

机构地区：[1]江南大学轻工过程先进控制教育部重点实验室,江苏无锡214122

出　　处：《传感器与微系统》2019年第1期107-109,共3页Transducer and Microsystem Technologies

基　　金：国家自然科学基金资助项目(61273131)

摘　　要：针对汽车气囊封条贴片表面缺陷检测效率低下的弊端,构建了一种基于视觉软件HALCON可对气囊封条贴片进行自动化检测的机器视觉系统。运用图像处理算法中形态学算子实现对封条表面0. 3 mm2以上的污渍、露白、残缺的检测。通过对1 000个样本的检测与分析,系统的正确率超过99%,误报率小于1%,漏报率为0,结果表明:设计的系统具有精度高、稳定性强的优点。Aiming at malpractice of the low efficiency in detecting defects of surface in automobile airbag seal,a machine vision system based on vision software HALCON which can automatically detect airbag seal patch is constructed. Use morphological operator in image processing algorithm to achieve detecting of stains,exposedwhite,incomplete whose size are more than 0. 3 mm2 on the surface of seal. Through detection and analysis on 1 000 samples,correct rate of the system is higher than 99 %,false alarm rate is less than 1 %,and missing report rate is zero. The results show that the system has the advantages of high precision and strong stability.

关 键 词：缺陷检测 形态学 模板匹配 气囊封条 机器视觉 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]