基于动态模糊聚类的织物织造疵点检测算法

 针对目前织物疵点检测算法普遍存在的适应性不强，实时性不高等问题，通过对织物织造疵点的深入分析，提出一种基于动态模糊聚类的织物织造疵点检测算法。该算法在对织物图像进行预处理之后，以织物图像的经纬向灰度均值投影为特征值，然后根据疵点区域灰度均值投影的畸变现象，利用动态模糊聚类算法分离出可能的疵点区域，最后设置合适的畸变密度和畸变度阈值对“伪疵点”区域实施有效过滤，以识别并定位疵点区域。实验证明，该算法可靠稳定，适应性强，并且具有较强的抗噪声干扰的能力。

Fabric weaving defect detection algorithm based on dynamic fuzzy clustering

A fabric weaving defect detection algorithm based on dynamic fuzzy clustering is proposed to solve problems such as poor adaptability and real-time in fabric defect detection algorithms. The gray average projection in the weft and warp direction is regarded as eigenvalue after images preprocessing. The suspicious defect region is separated with dynamic fuzzy clustering algorithm according to the projection distortion on the gray average in defect region, which is located and extracted from suspicious defect region by selecting proper thresholds of the distortion density and degree so that the pseudo defect region is filtered out effectively. The results of experiments show that this algorithm has the features of high reliability, strong adaptability and anti-noise-interference ability.