摘要
针对目前在印刷及包装行业板纸机械式计数存在的噪声大、效率低、对板纸有损伤等不足,设计了基于局部灰度特征的纸张计数算法并实验验证了该算法的计数效果。该方法从板纸图像的列灰度特征出发,绘制板纸列像素灰度值变化曲线,对列像素灰度连续做减法,以找到纸张与纸张间隙灰度变化的跳变点,对得到的灰度差曲线做二值化处理,最后根据所找到的跳变点进行计数,并依据纸张宽度剔除干扰的标注点,以使计数结果更加准确。实验结果表明:该方法对于纹理单一、灰度对比明显的纸张,其计数效果理想,准确率高,达9 7%以上;且由于该算法从灰度图像列灰度特征入手,对一维数据进行处理,因此运行速度快,效率高,能满足实时快速处理的要求。
Aimed at the shortcomings of noise, inefficiency and damage in mechanical counting in printing and packaging industry, the paper counting algorithm based on local gray feature was designed with the effect of the algorithm verified by experiment. Based on column gray feature, the method could draw column pixel gray value curve, then column pixel gray was subtracted continuously for finding the transition point of paper spacing variations. Gray difference curve was subjected to binary processing; finally counting based on transition point was achieved. By eliminating interference points according to the paper width, the counting results were made more accurate. The experimental results showed that for paper with single texture and obvious gray difference, the method of counting achieved the ideal effect with high accuracy of up to 97%. As the algorithm was based upon the column gray by processing one-dimensional data, the running speed was fast with high efficiency, which could meet the requirement of real-time processing.
出处
《包装学报》
2016年第4期24-29,共6页
Packaging Journal
基金
湖南省自然科学基金资助项目(2016JJ6034)
湖南省产业化培育基金资助项目(15CY003)
关键词
局部灰度特征
板纸计数
列灰度
纸张宽度
标注点
local gray feature
paper counting
gray column
width of the paper
label points
