中文摘要：

为实现经多次洗涤后织物色牢度等级评定方法的客观化，提出了一种基于数字图像处理的织物耐洗涤色牢度等级评定方法。该方法首先采用扫描仪对AATCC沾色灰卡进行扫描，并计算灰卡各等级的灰度值，然后对沾色灰卡各等级与其对应灰度值进行线性拟合，得到沾色等级-灰度值方程。在采用图像处理方法对织物色牢度等级评定时，织物易出现沾色不匀的现象，其对评级结果的准确性存在很大的影响，因此在将试样灰度值代入方程计算之前，需要对沾色区域进行分割，并对沾色区域灰度值进行计算。但是，直接对图像进行分割时，纹理信息会产生一定的影响。因此，在计算沾色区域灰度值前先利用傅里叶变换滤除织物纹理，然后对滤除纹理后的图像进行沾色区域分割，从而对沾色区域灰度值进行精确计算，提高沾色等级评价精度。

英文摘要：

In order to evaluate the color fastness grade after frequent laundering objectively, a method based on image analysis is proposed. Firstly, the images of different grades of the Gray Scale for Staining for AATCC are scanned to detect the gray value. Secondly, the equation of grade for the equation of grade for staining-gray-scale value are fitted by analyzing the relationship between the gray value of its corresponding grade. The staining of fabric is unevenness, which greatly affect the accuracy of the result, so before detect the gray value the image should be segmented with the gray value calculated. However, the texture of fabric will have the effect on image segmentation. In this article, the fast fourier transform was used to realize the filter of fabric texture and separate staining region, then accurately calculate to improve the color fastness grade evaluation.