纸币号码图像预处理及识别技术

Image pretreatment and recognition technology of banknote number

在不去除纸币号码图像背景的前提下,为了进一步提高纸币号码图像识别的工作效率和准确度,本文对纸币号码图像的识别技术做了相应研究。运用图像增强、图像灰度形态学平滑去噪、图像二值化和识别区域的定位等方法对图像进行预处理。以实验分析为基础,应用模板匹配算法和特征结构识别算法相结合来实现纸币号码识别。实验结果表明,该方法针对印度币和人民币号码图像的识别率、识别时间均达到实际使用要求,是能够满足实际需要的综合识别算法。

In order not to remove the background of banknote number image,and improve the efficiency and accuracy of recognition,the recognition technology of banknote image is researched. By using the image en-hancement,image gray-scale morphological smoothing denoising,binarization and positioning of identifying ares,methods image preprocessing is carried out. On the basis of experimental analysis,by using the template matching algorithm and the characteristics of the structure identification algorithm,the banknote recognition is realized. The experimental results show that the method for India and RMB number recognition rate and rec-ognition time can meet the requirements of actual use,which is a comprehensive recognition algorithm for ac-tual needs.