摘要
首先对钢帘线材图像进行预处理,以减小或消除噪声的影响,然后对钢帘线图像进行了缺陷特征分析,运用经典的边缘提取算法提取线材的面积与周长等几何特征,运用灰度共生矩阵来提取线材的纹理特征等来判断钢帘线表面是否有缩颈或耳子等缺陷.实验结果表明了该特征提取算法的有效性和实用性.
Feature analysis of surface defect is a key factor for the steel wire cord surface defects auto-detecting system on image processing to work perfectly and accurately. First of all, the images of steel wire cord have been pretreated in order to reduce or eliminate the influence of noise, then with the method of the classical edge detection algorithm and gray co-occurrence matrix, and the geometical features of the image,such as the size and wire perimeter, the texture features are picked up to decide whether the steel wire cord has defects on its surface. The Results of experiments show that the feature extraction algorithm is effective and practicable.
出处
《湖北工业大学学报》
2008年第2期26-29,共4页
Journal of Hubei University of Technology
关键词
钢帘线
缺陷检测
图像处理
纹理特征
几何特征
steel wire cord
defect detecting image processing texture feature geometical feature
