摘要
为实现机织物密度的自动测量,设计了简单的图像采集装置以获得高清的织物图像.利用MATLAB编写算法,对织物图像进行图像灰度化、直方图规定化、中值滤波、灰度增强、带通滤波等处理,从而获得织物经纬方向的信息,得出机织物的经纬密度.实验表明,本算法不仅适用于纯色织物,也适用于一些简单的印花织物,与人工检测结果相比,机器检测结果的相对误差均在2%以内,假设检验结果说明算法具有较高的可信度和稳定性.
For detecting fabric density automatically, a simple system for getting HD fabric image was built. MATLAB-based algorithms, gray level transformation, histogram specification, median filtering, gray level increase, and band-pass filter were used to deal with fabric image to get fabric density automatically. Experiments show that the algorithms are suited for plain fabrics and some simple printed fabrics; contrast experiments show the relative error between manual measurement and image measurement is less than 2%, and the hypothesis testing results demonstrate that the algorithms have high reliability and stability.
出处
《五邑大学学报（自然科学版）》
CAS
2015年第2期16-20,共5页
Journal of Wuyi University(Natural Science Edition)
