基于灰度值差分的纸张计数算法设计

Design of Paper Counting Algorithm Based on Gray Value Difference

针对实际生产中机械式纸张计数存在成本高、噪声大的缺点,采用基于灰度差分的纸张计数算法,并且实验验证了其计数的准确性。基于灰度差分的纸张计数算法将排列整齐的纸张的图像进行灰度化后,利用Gabor滤波器进行图像边缘增强,并利用图像灰度投影后的灰度差分,得到灰度像素曲线;根据波峰与波峰之间的间隔初步判断1张纸的厚度,排除干扰峰谷点后,准确定位出波峰或者波谷的的位置;根据标记的位置,实现对纸张进行计数。实验验证结果表明:该算法能较好地实现对纸张的计数,且简单易行,准确率在97%以上。

Aimed at the shortcoming of high cost and noise in mechanical paper counter, the paper counting algorithm based on gray value difference of paper was adopted and tested in order to resolve these issues from a visual way. After the image graying of neatly arranged paper, with the Gabor filter grayscale image edge enhancement, the gray pixel curve was obtained by using the gray value difference after gray projection. The thickness of a sheet of paper could be preliminarily judged based on the interval of peaks as the position of peak or valley could be accurately located after the exclusion of peak valley disturbance. Then paper counting was conducted according to the location markers. The counting results showed that this paper counting method could perform with simple operation, while the accuracy rate of paper counting achieved more than 97%.