钞币面值和真伪的快速识别算法的研究

根据从传感器采集来的信号的特点,提出了在验钞系统中进行鉴伪和面额区分的模式识别算法,并详细阐述了算法的整个实现过程。为了识别钞票面额,采用了信号模板匹配的方案。为了鉴别钞票的真伪,应用了统计模式识别技术,并给出了进行真伪钞判别的线性函数。该算法已经应用到嵌入式系统当中,结果表明:并具有快速,准确,有效等优点,并具有非常好的市场应用前景。

Study of Fast Recognition Algorithm for Banknote Denomination and Counterfeit Discriminating

In this paper,the algorithm of pattern recognition for banknote is given,based on the characteristic of signal from sensors and the whole implementation of this algorithm is introduced.In order to recognize the denomination of ifferent notes,a solution of template matching is applied to this system,and the technology of statistical pattern recognition is also used in this design in order to discriminate counterfeit from true notes,which depends on the linear function from collected samples.These algorithms are already applied to the embedded-system.A lot of experiments in practice indicate that the algorithms can distinguish between the true and false banknotes quickly,correctly and effectively.So the system designed on these algorithms must be widely applied to market in the future.