零件的机器视觉自动识别检测系统

Design on Parts' Automatic Detection and Identification System on Machine Vision

为了实现零件识别检测的自动化与零件生产的智能化,将机器视觉与运动控制理论结合,采用脉冲耦合式神经网络PCNN(pulse coupled neural network)边缘检测识别的方法,初步设计出一套机器视觉自动识别检测系统,对继电器盖等小型零件进行了检测识别实验。实验结果表明,该视觉检测识别自动化系统改善了零件检测自动化程度,提高了尺寸精度,优化了系统鲁棒性,最终实现了对零件实时筛选分类与零件的高效率生产。

In order to realize the automatic detection and identification of parts and intelligentize the manufacturing process, a preliminary design for automatic detection and identification system based on pulse coupled neural network （PCNN） is put forward by combining machine vision and motion control theory. Cor-responding experiments for relay cover have been carried out. It shows that the proposed system can increase the level of automatic detection. And it also reveals that the size precision of parts and the robustness of the system have been improved. Conclusion are drawn that the system can classify the parts in real time and in high productivity.