基于计算机图像处理技术的高速纸机纸病在线检测研究

Research on Online Detection of Paper Disease in High Speed Paper Machine Based on Computer Image Processing Technology

现代造纸工业向着生产自动化、设备大型化的方向飞速发展,人工的纸张缺陷检测方式已无法满足现实需求,研究基于计算机图像处理技术的纸病在线检测技术,对于提升纸病在线诊断效率和水平显得尤为重要。对此,简述了纸病在线检测系统的基本原理,搭建了实验平台,重点设计了纸病图像在线检测算法,采用中值滤波进行纸张图像预处理,基于差影算法和人工蜂群算法优化的二维阈值分割算法进行纸病图像检测,训练好分类器后对测试样本进行分类识别,结果表明,多种纸病的平均检测正确率在95%以上,可较好满足现实需求。

With the rapid development of modern paper industry towards the direction of production automation and large-scale equipment,the artificial paper defect detection method can no longer meet the practical needs.It is particularly important to study the online paper disease detection technology based on computer image processing technology to improve the efficiency and level of online paper disease diagnosis.In this paper,the basic principle of online paper disease detection system is described,the experimental platform is set up,the paper disease image online detection algorithm is mainly designed,the paper image is preprocessed by the median filter,the paper disease image detection is carried out by the two-dimensional threshold segmentation algorithm optimized by the difference shadow algorithm and artificial bee colony algorithm,and the test samples are classified and recognized after the classifier is trained.The results show that,the average detection accuracy of various paper diseases is more than 95%,which can better meet the practical demand.