基于OpenCV的火车票识别算法

The algorithm of train ticket identification based on OpenCV

针对火车站的机器验票不能验证持票人与证、票相一致的不足,设计了一种利用摄像头拍摄车票图像、基于OpenCV视觉库的火车票身份证号码和乘车日期识别算法.算法利用车票上的二维码角点位置定位相关字符区域,采用适用于不同亮度情况下的自适应阈值分割算法实现字符分割;身份证号码识别算法利用目标外轮廓特征提取字符目标,使用模板匹配法识别目标;日期识别算法采用漫水填充法提取字符目标,根据日期字符按模块排列的特征识别目标.试验结果证实该算法在保证字符识别正确率的同时可以满足系统实时性的要求,解决了人工验证旅客身份效率低的问题.

As railway station ticket checking machines are unable to verify the consistency of the ticket-holder, IDcard, and ticket, an OpenCV and camera image-based algorithm is developed for identifying train ticket, ID cardnumber, and boarding date. The preliminary experimental results are given and the method has a certain academicand application values. The algorithm uses QR code’s corner position to locate relating character, adopt adaptivethreshold segmentation algorithm which is suitable for different brightness to realize the character segmentation.The recognition algorithm of ID uses the contour of character to extract the characters, and it uses the templatematching method to identify the target. The recognition algorithm about boarding date uses the flood-fill methodaccording to the characteristics of which the modules are arranged. Experiment results prove that the algorithmguarantees the character recognition’s accuracy and meets the real-time requirement. It solves the low efficiency ofmanual validation of passenger identity.