| 摘要: |
| 自动纸币清分机是提高银行效率的一种重要工具。开发的一种基于图像处理技术的纸币清分机采用污渍识别算法，即将整张纸币的图像数据分成固定大小的组，求出每组像素灰度的均值和方差，并与预先设定的门限值进行比较，完成污溃属性的识别。该算法采用FPGA硬件实现，占用很少的系统处理时间，达到90％的识别率。 |
| 关键词: 图像处理 纸币清分机 污渍识别 均值 方差 算法 |
| DOI：10.3969/j.issn.1001-893X. |
| 投稿时间：2005-08-23修订日期：2005-12-14 |
| 基金项目:中国科学院资助项目;广东省深圳市科技局科研项目 |
|
| A Spots Recognition Algorithm and its Implementation |
|
| () |
| Abstract: |
| An auto notes discriminator, which is used to discern the notes attributes, such as denomination, direction, dilapidation, grades and spots etc, is an important tool to improve the efficiency in banks. A notes discriminator based on image processing technology has been developed. Spots recognition algorithm is adopted. That is, first, to seek the mean and variance value of a group of pixels and then compare with fixed thresholds. The spots recognition results can be got according to the comparative results. This algorithm is implemented by FPGA device. It only need a little processing time in the whole system and can achieve a 90% discrimination rate. |
| Key words: image processing notes discriminator spots recognition mean value variance value algorithm |
《电讯技术》编辑部 