基于机器视觉缺陷检测可重构系统的设计与实现

Design and implementation of reconfigurable system based on machine vision defect detection

设计一种能够应用于不同规格产品缺陷检测的软件,以满足中小企业对于不同产品质量检测需求。在分析传统缺陷检测系统技术的基础上,利用PYTHON语言,将阈值率定设置、ROI标识、算法优选、机器学习及持久化存储等工程设计为用于实验的设置程序。利用软件设计等方面的技巧,使用C++语言,根据预设图像处理算法和步骤,动态构建在线检测程序。使得在线检测程序职责单一化,实现不同产品缺陷在线检测程序。改变了传统的缺陷检测系统是针对某一产品进行硬编码的专用系统,降低了系统的开发和使用成本,对资金薄弱的中小代工企业进行生产技术改造,提高生产的柔性和自动化程度有一定的理论与实践意义。

Design a software that can be applied to different product defect detection to meet the demand of small and medium-sized enterprises for different product quality. On the basis of analyzing the traditional defect detection system technology, using the PYTHON language, set the threshold rate ROI identification algorithm optimization of machine learning and persistent storage and other projects designed to be used in the experiment setup using skills such as the software design, using c ＋ ＋ language, according to the default image processing algorithm and steps, dynamically build online test procedures. Makes the online test procedure duties simplification and implement different online product defect detection programs changed the traditional defect detection system is a product for hard coded dedicated system, reduces the development and use of system cost, the capital of weak production technology innovation of small and medium sized foundry enterprises, improve the flexibility and automation of production has a certain theoretical and practical significance.