人民币纸币红外、紫外特征检测方法

根据国家2010年公布的《鉴别仪通用技术条件》对纸币图像分析鉴别能力要求的提升,在传统点钞机白光图像识别的基础上增加了纸币红外、紫外图像鉴伪功能.目前国内主流点钞机出钞速度为20张/秒左右,为了与这个速度匹配的同时又保证识别准确率,提出了接触式图像传感器加DSP加可编程逻辑器件CPLD 的高速图像采集方法.大量实验表明,同时完整地采集到三幅高质量的纸币可见光图像及红外、紫外特征图像共需40.5毫秒,满足主流点钞机单张走纸50毫秒的速度要求.

RMB Banknotes Infrared, Ultraviolet Characteristics Detection Method

According to the requirements for image analysis and identification ability from state in 2010 published "identifying instrument general technical conditions", the paper currency image authentication function is added in the traditional counting money machine. At present, domestic mainstream counting machine, banknote at a speed of 20 frames per second or so, in order to match the speed and ensure the accuracy, a high speed image acquisition method based on contact image sensor and DSP and programmable logic device CPLD is proposed. A large number of experiments show that at the same time to complete collection three high quality images of the visible light, infrared image and ultraviolet image just need 40.5ms, to meet the requirements for current market currency count machine sheet of paper 50 millisecond speed.