摘要
纸币粘连是导致点钞机计数失败的主要原因,粘连币可以细分为:纸币部分重叠并相连的连张币和纸币边缘完全重叠的重张币。运用速度传感器并配合左右计数红外传感器,测量纸币的宽度并进行纠偏、补孔运算,不仅能检测出连张币,还能检测残缺、裂口、卷边纸币,检测精度不低于2mm;同时通过复用左右两对红外计数传感器作为纸质传感器,并将透射过纸币的红外信号进行放大并连续进行A/D转换,应用滑动窗口实时建立纸币纸质特征隶属函数,进行纸币重张模糊识别同时又能检测出粘补纸币。运用上述方法设计的LD-50D型点钞机能自动清点美元、欧元、港币、卢布、澳元以及罗马尼亚等国的不同纸币,错点率指标优于GB16999-1997规定的10-5。
The main reason which causes the counting failure of bank machines is stuck banknotes. Stuck banknotes are subdivided into two types: continuous banknotes and double banknotes. The speed sensor together with the infrared sensors can measure the width of banknotes, correct errors caused by banknote slanting, and print holes. So the machine can not only check out continuous banknotes but also find out misshapen chasm banknotes. The accuracy of measurement is not lower than 2 mm. At the same time, the infrared signals through the banknotes are magnified and converted into digital signals by ADC.The membership function of banknotes is constructed on real time by sliding windows so that double banknotes can be checked out. Designed in this way, the machine can count banknotes of any countries such as USD, EURO, HKD, RUB, and AUD without manual selecting , its accuracy being better than 10^(-5) designated in GB16999-1997.
出处
《浙江工业大学学报》
CAS
2004年第2期220-224,共5页
Journal of Zhejiang University of Technology
关键词
粘连币
模糊识别
动态鉴别
隶属函数
红外传感器
点钞机
stuck banknote
dynamic discrimination
fuzzy identification
membership function
infrared sensor
