摘要
开发了一种快速的纸币号自动识别系统,能在捆钞的同时自动识别并记录下通过捆钞机上的纸币号码。该系统首先通过CCD摄像头采集24位彩色图像,经过逻辑脊谷判断法分割、投影法定位、结构法识别几个过程输出号码串。实验结果表明,该系统组织设计的逻辑脊谷判断算法有效提取了号码图像,穿越号码体距离法定位号码准确,提出的号码识别系统取得了较高的识别率。
An automatic recognition system for paper currency numbers is developed. The paper currency number is recognized and recorded with binding by the system. Firstly, 24 bits color image is gathered through CCD camera, then outputs number clusters through process of segmentation by gray ridge-vale algorithm, orientation by projection, and character recognition by structure-analyzing algorithm. Experimental results demonstrate the proposed algorithm of gray ridge-vale can extract number image effectively, the distance of traversing number body is accuracy, the recognition system achieves a high recognition rate and fast recognition speed.
出处
《计算机工程》
EI
CAS
CSCD
北大核心
2005年第24期153-155,共3页
Computer Engineering
基金
辽宁省教育厅高校科研资助项目(2004C069)
关键词
号码识别
纸币号码
逻辑判断
图像分割
Number recognition
Paper currency numbers
Logic judgment
Image segmentation
