啤酒瓶在线模号识别系统研究

为了对啤酒瓶进行准确快速的分类,提高其生产效率,设计并实现了啤酒瓶在线模号识别系统;提出采用基于光反射的非接触式检测方法提取模点信息,并且比较了两种不同的光接收方式,提出了插入式模号识别算法,该算法具有识别效率高的特点;着重介绍了检测系统的结构组成,该系统具有准确度高、稳定性好、抗抖动等优点,与采用CCD方案识别玻璃瓶文字相比此系统对硬件要求较低,具有很高的性能价格比;在这种系统结构上结合插入式模号识别算法可以提高识别速度和识别率,能成功地应用于生产实践中。

Research on the On-line Detection System of Beer Bottle's Model

In order to improve the production of beer bottles effectively,an on-line mould number detection system is designed and realized to classify these bottles.A non-contact detect method based on light reflect was developed for obtaining mould number information,and also compared the two different light receive models;and also give the plug-in recognition algorithm,which was effective.The structure of the detection system is introduced,this system has a good veracity,stability and regardless of dithering.Compared with using CCD to identify bottle text,this system is of high cost performance because it has low requirement to computer hardware.Use the plug-in recognition algorithm on this system structure can enhance the recognition speed and the recognition rate.And this on-line detection system has been well applied to production field.