基于LCSM算法的卷烟包装纸缺陷检测研究  

Cigarette Packaging Paper Defect Detection Research Based on LCSM Algorithm

作　　者：高云[1] 郝冠刚 王冠 陆伟 邓来宏 石磊 GAO Yun;HAO Guangang;WANG Guan;LU Wei;DENG Laihong;SHI Lei(China Tobacco Shandong Industrial Co.,Ltd.,Jinan 250014,China;Shandong Tobacco Quality Supervision and Testing Station,Jinan 250101,China)

机构地区：[1]山东中烟工业有限责任公司,山东济南250014 [2]山东省烟草质量监督检测站,山东济南250101

出　　处：《造纸科学与技术》2025年第1期12-16,共5页Paper Science And Technology

基　　金：山东中烟工业有限责任公司科技项目(2024370000340122)。

摘　　要：受材料特性以及制造工艺的影响,卷烟包装纸的表面缺陷通常与正常纹理交织在一起,导致难以对其进行精准定位和识别。为提高卷烟包装纸质量控制水平,现提出基于LCSM算法的卷烟包装纸缺陷检测研究。通过对灰度变换函数进行定义,在指定区间内对原始灰度图像像素灰度值进行调整,并结合空间域和灰度域信息,对其进行非线性滤波处理。结合LCSM算法,通过构建高斯金字塔,计算图像梯度,得到局部特征描述子。以余弦相似度作为匹配指标,通过特征描述子匹配找到待测图像与参考图像中的对应点,并计算这些点周围的局部对比度。最后结合差异度阈值,对区域是否存在缺陷进行判断。在实验中,对提出的方法进行了检测精度的检验。最终的测试结果表明,采用提出的方法对卷烟包装纸进行缺陷检测时,算法在不同清晰度环境下均能保持较高的缺陷区域重叠度,具有较为理想的检测效果。Influenced by the material properties and manufacturing process,the surface defects of cigarette wrapping paper are usually intertwined with the normal texture,which makes it difficult to accurately locate and identify them.In order to improve the quality control level of cigarette wrapping paper,the defect detection research of cigarette wrapping paper based on LCSM algorithm is proposed.By defining the grayscale transformation function,the original grayscale image pixel grayscale value is adjusted in the specified interval,and combined with the spatial domain and grayscale domain information,it is subjected to nonlinear filtering.Combined with the LCSM algorithm,the local feature descriptors are obtained by constructing a Gaussian pyramid and calculating the image gradient.Taking the cosine similarity as the matching index,the corresponding points in the to-be-tested image and the reference image are found by feature descriptor matching,and the local contrast around these points is calculated.Finally,combined with the difference threshold,the region is judged whether there are defects or not.In the experiments,the detection accuracy of the proposed method is tested.The final test results show that the algorithm can maintain a high degree of overlap of defective regions under different clarity environments when the proposed method is used for defect detection on cigarette wrappers,which has a more ideal detection effect.

关 键 词：卷烟包装 缺陷检测 检测精度 LCSM 局域特征相似性 

分 类 号：TS76[轻工技术与工程—制浆造纸工程]