摘要
棉、毛纤维广泛应用于混纺织物中,如何在混纺织物中快速地识别出该类纤维具有重要的实用价值。文章提出了一种基于SIFT(尺度不变特征变换)和KNN(K最近邻)图像处理算法的识别方法,以棉、毛纤维的电镜图像作为样本,进行图像预处理,然后使用SIFT算法进行特征提取,生成特征描述子,最后采用KNN分类算法进行匹配、分类,得到被测纤维图像的种类,实现棉、毛纤维的快速自动识别。实验结果表明:该方法无需大量数据样本,识别率超过95%,且具有检测速度快、稳定等优点,实用性强。
Cotton and wool fibers are widely used in blended fabric,how to quickly identify these fibers in fabric has important practical value.A recognition method based on SIFT(Scale Invariant Feature Transformation)and KNN(K Nearest Neighbor)image processing algorithm was proposed.The electron microscope image of cotton and wool fiber were taken as a sample,image preprocessing was carried out,and then SIFT(Scale Invariant Feature Transformation)algorithm was used to extract features and generate feature descriptors.Finally,KNN(K Nearest Neighbor)classification algorithm was used for matching and classification to obtain the types of fiber image to realize fast and automatic recognition of cotton and wool fiber.Experimental results show that this method does not need a large number of data samples,the recognition rate is more than 95%,with the advantages of fast detection speed,stability and strong practicability.
作者
游小荣
李淑芳
YOU Xiaorong;LI Shufang(Changzhou Textile Garment Institute,Changzhou,Jiangsu 213164,China;Changzhou Key Laboratory of Eco-Textile Technology,Changzhou,Jiangsu 213164,China)
出处
《毛纺科技》
CAS
北大核心
2021年第8期91-94,共4页
Wool Textile Journal
基金
江苏省高等学校大学生实践创新训练项目(201912807005Y)。
关键词
棉纤维
毛纤维
混纺
自动识别
SIFT
KNN
cotton fiber
wool fiber
blending
automatic identification
SIFT
KNN
