纱线扭结分布特征的计算机自动识别方法

Computerized characterization of the yarn snarling distribution

纱线单位长度上的湿扭结个数是评价和衡量纱线残余扭矩的一项重要间接指标。结合数字图像和信号处理技术,提出了一种基于纱线图像的扭结分布特征的自动识别方法。该方法首先提取纱线的扭结特征曲线,进而利用信号处理的手段实现了以若干高斯信号来自动匹配和识别纱线扭结特征曲线。实验结果表明,该方法可以自动准确地计算出纱线湿扭结个数,同时可获得纱线扭结高度及宽度2个几何参数。

The number of yarn wet snarling per unit length is an important indirect index for the evaluation of yarn residual torque. A computerized characterization method for the yarn snarling distribution was proposed by applying the digital image and signal processing techniques to the actual yarn snarling images. The yarn snarling characteristic curve is firstly extracted, then several Gauss signals would be calculated to match with the yarn snarling characteristic curve. Experimental results revealed that the number of yarn wet snarling can be automatically calculated with a high accuracy by the proposed method, besides, the height and width values of each snarling can also be obtained.