基于边界跟踪测量麻纤维横截面参数的算法研究与应用

针对在同一纤维横截面图像中难以同时测量多根纤维参数的问题,提出了一种基于边界跟踪测量麻纤维横截面参数的算法。首先利用边缘提取算法提取麻纤维的边缘,通过修改边缘提取算法的敏感度阈值获取合适的边缘检测图像,然后采用边界跟踪算法对边缘检测图像中的麻纤维进行识别和标记,利用循环标记出图像中所有的麻纤维并保存边界跟踪的路径点坐标,最后利用算法测量麻纤维的周长、面积和圆度。实验结果表明:该方法能够同时测量多根麻纤维的横截面参数;根据用于测试的标准圆的数据可知,本算法测得的麻纤维的周长偏差大约为3%,面积偏差大约为4%,数据误差较小。

Research and application of algorithm for measuring hemp fiber cross-sectional parameters based on boundary tracking

It is difficult to simultaneously measure the parameters of multiple fibers in the cross-section image of fibers.To solve this problem,an algorithm based on boundary tracking for measuring the cross-section parameters of hemp fibers was proposed.After the edges of hemp fibers were extracted by an edge extraction algorithm,an appropriate edge detection image was obtained by modifying sensitivity threshold of the edge extraction algorithm.The hemp fibers were then identified and marked by a boundary tracking algorithm.All hemp fibers in the image were marked by cyclic marking and path point coordinates of boundary tracking were preserved.Finally,the algorithm was used to measure the perimeter,area and roundness of hemp fibers.The experimental results show that the method can simultaneously measure cross-sectional parameters of several hemp fibers.According to the data of the standard circle used for testing,the circumference deviation and area deviation of hemp fibers measured by this algorithm are about 3%and 4%respectively,and the data error is small.