基于光电技术的棉纤维杂质快速测定

针对棉纤维杂质人工检测耗时长、劳动强度大等问题,提出一种基于光电技术的棉纤维杂质机器视觉测定方法。首先通过改变平滑模板窗口尺寸对棉纤维图像自适应中值滤波优化,然后采用Otsu算法自适应选取分割阈值,并遍历整个图像,进行边缘断裂端点的连接、边缘生长与连接,搭建光电检测系统原型,实现了棉纤维杂质的快速测定。实验结果表明:该方法能够有效去除图像噪声,避免伪边缘的产生,检测到的杂质边缘清晰完整。光电检测耗时较GB/T 6499—2012《原棉含杂率试验方法》降低95.1%,结果间误差为2.35%,杂质粒数与杂质面积比参数结果误差均在2.1%以内,能够有效用于棉纤维杂质的快速准确测定。

Rapid measurement of cotton fiber impurity content based on photoelectric detection technology

Aiming at the problems such as long time-consuming and labor-intensive manual detection of cotton fiber impurities,a machine vision measurement method for cotton fiber impurities based on photoelectric technology was proposed.First,the cotton fiber image was adaptively optimized by changing the smooth template window size.Then the Otsu algorithm was adaptively used to select the segmentation threshold and traverse the entire image to connect the edge break endpoints,edge growth and connection,and build a prototype of a photoelectric detection system to realize the rapid determination of cotton fiber impurities.The experimental results show that the method in this paper can effectively remove image noise,avoid the generation of false edges,and the edges of the detected impurities are clear and complete.At the same time,the photoelectric detection time is 95.1%lower than that of the national standard GB/T 6499—2012"Test method for percentage of trash content in raw cotton"method,and the error between the results of the photoelectric detection and the national method is 2.35%,the parameter error of the number of impurities and the area ratio of impurities is within 2.1%,which can be effectively used for the rapid and accurate determination of cotton fiber impurities.