摘要
本文引入多尺度二维小波分析织物的表面折皱,织物图像首先经小波变换分解成低频和高频两部分,再从中提取四种表征织物折皱的特征参数,并计算特征参数与不同折皱等级的相关系数,实验表明,四种特征参数均可从不同侧面反映织物的折皱程度,特征参数与织物的折皱程度基本呈线性相关,最后采用分形理论与该方法进行对比,以验证此方法的可行性。
In this paper, Multi-Scale two-dimension wavelet is imported to analysis fabric surface wrinkle. Firstly, Fabric image is decomposed into two parts through wavelet transform, one is low frequency, and the other is high frequency. Then, four kinds of parameter of fabric wrinkle are extracted from high frequency part, and calculated the coefficient between feature parameters and different wrinkle grades. The result shows that four feature parameters can reflect the fabric wrinkle degree from different sides; it has closely linear correlation between feature parameters and fabric wrinkle degree, Finally, fractal theory is applied to compare with the method mentioned in this paper, which will verify the feasibility of this method.
出处
《苏州大学学报（工科版）》
CAS
2004年第2期18-22,共5页
Journal of Soochow University Engineering Science Edition (Bimonthly)
基金
高等学校博士学科点专项科研基金资助课题。
