基于机器视觉的工业炸药包装缺陷在线检测系统  

An On-line Packing Inspection System of Industrial Explosives Based on Machine Vision

作　　者：黄志平[1,2] 许亮[3,2] 

机构地区：[1]广东省四〇一厂,广东梅州514795 [2]广东振声科技股份有限公司,广东梅州514795 [3]广东工业大学自动化学院,广东广州510006

出　　处：《爆破器材》2015年第3期43-47,共5页Explosive Materials

基　　金：国家自然科学基金(21376091);广东省重大科技专项(2012A080104012)

摘　　要：针对炸药药卷包装缺陷,开发了药卷包装缺陷在线视觉检测系统。对系统结构、视觉成像系统进行了描述,并详细介绍了检测方法。采用该方法对获取的药卷图像预处理,并计算药卷轮廓特征,将得到的药卷特征参数与预先建立的药卷标准模板进行对比。试验和应用表明,该方法对3类常见包装缺陷检测准确率不低于96%,所开发的系统可以快速、准确地检测药卷缺陷,能满足实际炸药包装生产过程在线检测的需要。An on-line inspecting system for packing defection of explosives was developed. The system structure,the vision system,and the inspecting method were presented. The acquired images of explosive cartridges were preprocessed using the proposed method,and the outline features were extracted from these images. Furthermore,the feature parameters obtained were compared to the template base pre-defined. The experiments and applications have show that the proposed approach can detect the common three packing defects and the inspecting accuracy reaches no less than 96%. The system can quickly and accurately detect the flaws and satisfy the requirements on the real production line.

关 键 词：工业炸药 包装缺陷 机器视觉 在线检测 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置] TD235.21[自动化与计算机技术—控制科学与工程]