列车车号视频识别系统  

Train Character Recognition System Based on Video Processing

作　　者：王剑[1] 赵歆波[1] 葛莉[1] 牛智慧[1] 

机构地区：[1]西北工业大学计算机学院,陕西西安710129

出　　处：《计算机技术与发展》2015年第3期28-31,共4页Computer Technology and Development

基　　金：国家自然科学基金资助项目(61171156;60872144)

摘　　要：数字图像识别技术正成为当前全球的热门话题,将图像识别技术成功地运用到实践中意义重大。文中针对现有货运列车车号识别系统的不足,提出了基于计算机视觉技术的货运列车车号识别的新方法。给出了系统的硬件组成与软件总体结构,说明了硬件系统设计和设备选型的方法;阐述了车号区域定位、车号字符分割与车号识别等关键技术的实现方法。软件模型由C#编程实现,实验结果表明该系统能够自动判断火车到达与离开,能采集清晰的图像,并实现列车车号的自动识别,准确地输出车号识别结果。Digital image recognition technology is becoming a hot topic in the world,the image recognition technology successfully applied in practice is of great significance. Aiming at the shortage of current freight train number recognition system,a new recognition method for freight train number is proposed based on computer vision technology. Firstly,the hardware components and overall structure of software are proposed. Then the design of the hardware system and the selection of the hardware devices type are described. After that,key algorithms,such as license plate area location,license plate segmentation and license plate recognition,are presented. The software system was developed with C#. At last,the experimental results show that the system can robustly capture clear image. Based on these images,automatically judge the train arrival or not. The license plate are precisely recognized,and recognition results are output in real time.

关 键 词：列车车号 车号区域定位 车号字符分割 车号识别 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]