Study Of Image Recognition Method On Paper Currency Numbers

The numbers on the paper currency is the amount identifier of the paper currency issuance, each sheet has its own numbers and the same numbers haven't be used twice, so the numbers on it can identify paper currency's identity. Developed a aptitude system by computer vision for paper currency numbers recognition, the number is recognized and recorded with binding by the system, then the organization of the paper currency could be easy. So the automatic register system has utilize tremendous and amplitude application foreground. In this paper, the automatic recognition system for paper currency numbers was built by the technique of image processing and pattern recognition.Firstly, compared the camera of CCD and CIS, a kind of Panasonic CCD camera was chosea The image acquire hardware was composed of CCD camera, image acquisition card and computer. Used the acquisition card DDL, the acquire program was programmed. The image gotten by camera has many noises. Through the experiment of average filter and median filter, the algorithm of Gauss was adopted to dispel the noises.Compared with the prevalence algorithm of segmentation, the experiment result show they are not content the system target in effect, speed and complexity. The logical judgment about ridge-valley algorithm was presented based on the character of number line's gray value is smaller than the background's gray value around them. The gray level of four directions was calculated. If one of the directions content the qualification, it means the gray level at the valley location can be judged as number line dot. Then the image was processed by iterative method binarization and morphology close operation.The noises was wipe off by calculate the areas and connect the breakpoints. In level location, the gray projection method was adopted. In perpendicular orientation, an improved method of distance of traversing number body was adopted. Then every number has a notable peak, and the two or more peaks made by projection directly is avoided.In recognition algorithm, the simple template matching has too much calculator, and it usually takes a long time. A new structure-analyzing algorithm was be put forward by calculate the transfer of the number body times. The algorithm is simply and fast enough to content the processing system.Take the RMB as an example, 300 sheets of 100 par value was random chose, the recognition ratio is 98 percent.