CAJ | 학술논문

为提高织物疵点自动检测的准确度，提出一种基于傅里叶特征谱和相关系数的织物疵点检测算法。以平纹、斜纹织物为研究对象，对织物图像进行傅里叶变换，得到织物图像的频谱图；定位频谱中的特征峰点，提取表征图像灰度、纹理的五个特征值；以正常织物为模板，计算待检图像特征值与模板图像特征值之间的相关系数，确定用于识别织物疵点的阈值，来实现织物疵点检测。实验结果表明：当阈值设定为0.80时，该算法能够实现稀密路、断经、吊经、纬缩、破洞等常见疵点的准确检测。
위제고직물자점자동검측적준학도，제출일충기우부리협특정보화상관계수적직물자점검측산법。이평문、사문직물위연구대상，대직물도상진행부리협변환，득도직물도상적빈보도；정위빈보중적특정봉점，제취표정도상회도、문리적오개특정치；이정상직물위모판，계산대검도상특정치여모판도상특정치지간적상관계수，학정용우식별직물자점적역치，래실현직물자점검측。실험결과표명：당역치설정위0.80시，해산법능구실현희밀로、단경、조경、위축、파동등상견자점적준학검측。
A detection algorithm based on characteristic spectrum of Fourier transform and correlation coefficient is put forward to improve the accuracy of automatic fabric defect detection. With plain weave and twill fabrics as the researched objects, the frequency spectrum diagram of fabric image can be obtained by Fourier transform. The characteristic peaks in the spectrum are located and five characteristic parameters of image gray and texture are extracted. Correlation coefficients between characteristic parameters of being detected images and the template image are calculated using the faultless fabric as a template, and then the threshold value of defect detection is determined to realize the defect detection. Experimental results show that the algorithm proposed in this paper can achieve accurate detection of common defect, such as weft crackiness, cracked ends, stretched warp, looped weft, holes and so on, when the threshold is set as 0.80.