摘要
纸币扎把的识别是纸币处理一个关键环节.目前识别纸币扎把的方法有图像颜色识别法和图像噪声边沿法,两种方法都需要保证扎把颜色与背景色有较大的差异,而且计算复杂.提出了图像分块的图像熵计算方式来完成纸币捆中的扎把区域识别的新思路,放宽了条件,提高了识别速度.
Identification of banknotes tie is a key part of the paper money processing.The way to the image color recognition method and image noise edge method need to ensure that the tie color and background color are quite different,and computational complexity.Put forward new idea which recognize the area of paper tape in the bundle of the paper money by image entropy calculation of image sub-block.It relaxes conditions and improves the recognition speed.
出处
《微电子学与计算机》
CSCD
北大核心
2012年第11期149-151,156,共4页
Microelectronics & Computer
基金
河北省科学技术研究与发展计划项目(114572126)
关键词
纸币扎把
图像分块
图像识别
图像熵
banknotes tie
image block
image recognition
image entropy
