一种用于工件3D条码识别的图象综合预处理方法

A 3D Barcode Preprocessing Scheme Based on Image Recognition

针对汽车零件金属蜂窝载体催化剂在线质量自动检测的需要 ,根据工件表面光学特性差、检测系统从高温处理前后的工件上所获信号的差别及所受干扰都很大的实际情况 ,利用条码结构自身的特点和相关性 ,提出一种 3 D条码图象综合预处理方法。即用配有辅助光源的工业摄像机摄取 3 D条码图象 ,经直方图均衡化、图象投影变换、自适应阈值法条空信号提取、条空信息真伪校正等处理后 ,按国标译码而最终得到工件编号。应用该方法对 1 0 0幅实际的工件条码图象进行实验 ,识别准确率达到 97% 。

In this paper,a novel 3D barcode pre processing method is suggested with an aim at the needs of online quality auto detection of the metallic honeycomb catalyst of automobil parts, based on the optical performance difference of work piece surface information difference obtained from the work pieces before and after being treated by hightemperature in the monitoring or detecting system and using the charecteristics and correlation of barcode itself. An industrial camera equiped with an auxiliary illumination takes 3D barcode images. They are processed via the histogram equalization, image projection transformation, the extraction of bar and space information by self adaptive threshold method and correction of bar and space information truth and false, and then,the serial number of working pieces are decoded in accordance with the national standard. This method is used to experiment 100 barcode actual working images,with the accurate rate of recognition outcomes reaching 97%,whereby indicating the availability of this method.