基于双端须丛试样的棉毛纤维长度频率分布测量

Length-Frequency Distribution Testing of Cotton and Wool Based on Dual-Beard Sample

基于新型双端须丛试样,利用图像处理方法可快速低成本地检测棉、毛等纤维的长度,但存在计算指标较少的问题。为此提出一种新算法,即把测得的双端须丛相对线密度曲线(类似纤维长度分布曲线)转换成质量加权长度频率直方图,同时避开可能放大误差的微分过程。该算法可通过对相对线密度曲线上一系列数据点进行简单加减运算,获得各个长度组纤维的质量分数,并根据定义计算任意长度指标。基于24种棉纤维试样和12种毛纤维试样的实际检测显示,计算出的直方图和几种长度指标值与AFIS、Almeter基准方法的检测结果高度一致,表明该方法能够利用双端须丛试样测得全面、准确的纤维长度信息。

The novel sample of dual-beard can be used to measure the length of cotton and wool fibers in an efficient and economical way by image processing method,but there are only a few available parameters.Therefore,a new calculation method is proposed to convert the relative linear density curve of dual-beard（similar to fibrogram）into mass-weighted length histogram and avoid the process of amplifying the error.Simple computations of addition and subtraction are applied to coordinate values of the curve’s plots.Thus the method can contribute to obtain the mass ratio of each length group and be further used to calculate any length parameter.Practical tests based on 24 cotton fiber samples and 12 wool fiber samples show that the calculated histograms and length parameter values have high agreement with those from reference methods like AFIS and Almeter.The result proves that the new calculation method is effective for providing comprehensive and accurate fiber length information with dual-beardsample.