正在加载图片...
第8卷第3期 智能系统学报 Vol.8 No.3 2013年6月 CAAI Transactions on Intelligent Systems Jum.2013 D0I:10.3969/i.issn.1673-4785.201211014 网络出版地址:htp:/www.cnki.net/kcms/detail/23.1538.TP.20130515.0942.011.html 复 Contourlet和各向异性扩散的织物疵点图像降噪 吴一全123,万红1,叶志龙 (1.南京航空航天大学电子信息工程学院,江苏南京210016:2.东华大学纺织面料技术教育部重点实验室,上海 201620:3.浙江理工大学先进纺织材料与制备技术教育部重点实验室,浙江杭州310018) 摘要:图像降噪是织物疵点自动检测的首要步骤,其效果直接影响后续的图像分割、特征提取及识别结果.为进一 步改善降噪性能,提出了一种基于复Contourlet变换和各向异性扩散的织物疵点图像降噪方法.首先通过复Contour- let变换将织物疵点图像分解成低频和高频分量:然后分别利用P Laplace算子和Catte PM模型进行相应的扩散:最 后经复Contourlet逆变换重构疵点图像.大量实验结果表明,与小波阈值收缩和全变差扩散的混合方法、小波与PM 模型扩散相结合的方法、Contourlet结合全变差和自适应对比度扩散的方法、非下采样Contourlet结合非线性扩散的 方法相比,所提出的方法在主观视觉效果和客观定量评价指标上都有了较大的提高,更好地保留了织物图像的纹理 细节信息,说明了其降噪能力更强,能够有效地抑制噪声 关键词:织物疵点检测:织物疵点图像:图像降噪:复Contourlet变换:各向异性扩散:P Laplace算子:Catte PM模型 中图分类号:TP391.4:TS103.7文献标志码:A文章编号:1673-4785(2013)03-0214-06 中文引用格式:吴一全,万红,叶志龙.复Contourle和各向异性扩散的织物疵点图像降噪[J].智能系统学报,2013,8(3):214-219. 英文引用格式:WU Yiquan,WAN Hong,YE Zhilong.Fabric defect image noise reduction based on complex contourlet transform and anisotropic diffusion[J].CAAI Transactions on Intelligent Systems,2013,8(3):214-219. Fabric defect image noise reduction based on complex contourlet transform and anisotropic diffusion WU Yiquan'2.3,WAN Hong',YE Zhilong' (1.College of Electronic and Information Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China; 2.Key Laboratory of Textile Science Technology,Ministry of Education,Donghua University,Shanghai 201620,China;3.Key La- boratory of Advanced Textile Materials and Manufacturing Technology,Ministry of Education,Zhejiang Sci-Tech University,Hangzhou 310018,China) Abstract:Image noise reduction has been identified as the first step of automatic fabric defect detection.Its effect has a direct impact on subsequent image segmentation,feature extraction and the final recognition result.To im- prove the performance of noise reduction,a method of noise reduction for a fabric defect image based on complex contourlet transform and anisotropic diffusion has been proposed.Firstly,a fabric defect image was decomposed into low-frequency and high-frequency components through complex contourlet transform.Next,a P_Laplace operator and Catte_PM model were used to diffuse low-frequency and high-frequency components,respectively.Finally,the defect image was reconstructed by inverse complex contourlet transform.A large number of experimental results in- dicate that,compared with the hybrid method of wavelet threshold shrinkage with total variation diffusion,the meth- od combining the wavelet with PM model diffusion,the method combining contourlet with total variation and adap- tive contrast diffusion,and the method combining nonsubsampled contourlet with nonlinear diffusion,the proposed method has great improvement in both subjective visual effect and objective quantitative evaluation indicators, which can preserve the texture and details of fabric image better.The proposed method has stronger noise reduction capabilities and can suppress noise effectively. Keywords:fabric defect detection;fabric defect image;image noise reduction;complex contourlet transform;ani- sotropic diffusion;P_Laplace operator;Catte_PM model 织物疵点是纺织品生产过程中影响产品质量的 重要因素之一.传统的人工检测方法速度慢,劳动强 度大,易造成误检和漏检.基于图像处理技术的自动 收稿日期:2012-11-08.网络出版日期:2013-05-15. 检测系统可以极大地提高织物疵点的检测效率和精 基金项目:国家自然科学基金资助项目(60872065):纺织面料技术教 度,减少人为因素制约,大大提高了纺织品生产流程 育部重点实验室开放基金资助项目(P1111):先进纺织材 料与制备技术教育部重点实验室开放课题研究基金资助 中的自动化程度.织物疵点图像获取和传输过程中 项目(2010001):江苏高校优势学科建设工程资助项目. 通信作者:吴一全.E-mail:nuaaimage@163.com. 常受到高斯噪声和微量椒盐噪声的污染),致使图第 8 卷第 3 期 智 能 系 统 学 报 Vol.8 №.3 2013 年 6 月 CAAI Transactions on Intelligent Systems Jun. 2013 DOI:10.3969 / j.issn.1673⁃4785.201211014 网络出版地址:http: / / www.cnki.net / kcms/ detail / 23.1538.TP.20130515.0942.011.html 复 Contourlet 和各向异性扩散的织物疵点图像降噪 吴一全1,2,3 ,万红1 ,叶志龙1 (1.南京航空航天大学 电子信息工程学院,江苏 南京 210016; 2.东华大学 纺织面料技术教育部重点实验室,上海 201620; 3.浙江理工大学 先进纺织材料与制备技术教育部重点实验室,浙江 杭州 310018) 摘 要:图像降噪是织物疵点自动检测的首要步骤,其效果直接影响后续的图像分割、特征提取及识别结果.为进一 步改善降噪性能,提出了一种基于复 Contourlet 变换和各向异性扩散的织物疵点图像降噪方法.首先通过复 Contour⁃ let 变换将织物疵点图像分解成低频和高频分量;然后分别利用 P_Laplace 算子和 Catte_PM 模型进行相应的扩散;最 后经复 Contourlet 逆变换重构疵点图像.大量实验结果表明,与小波阈值收缩和全变差扩散的混合方法、小波与 PM 模型扩散相结合的方法、Contourlet 结合全变差和自适应对比度扩散的方法、非下采样 Contourlet 结合非线性扩散的 方法相比,所提出的方法在主观视觉效果和客观定量评价指标上都有了较大的提高,更好地保留了织物图像的纹理 细节信息,说明了其降噪能力更强,能够有效地抑制噪声. 关键词:织物疵点检测;织物疵点图像;图像降噪;复 Contourlet 变换;各向异性扩散;P_Laplace 算子;Catte_PM 模型 中图分类号: TP391.4;TS103.7 文献标志码:A 文章编号:1673⁃4785(2013)03⁃0214⁃06 中文引用格式:吴一全,万红,叶志龙.复 Contourlet 和各向异性扩散的织物疵点图像降噪[J].智能系统学报, 2013, 8(3): 214⁃219. 英文引用格式:WU Yiquan, WAN Hong, YE Zhilong. Fabric defect image noise reduction based on complex contourlet transform and anisotropic diffusion[J]. CAAI Transactions on Intelligent Systems, 2013, 8(3): 214⁃219. Fabric defect image noise reduction based on complex contourlet transform and anisotropic diffusion WU Yiquan 1,2,3 , WAN Hong 1 , YE Zhilong 1 (1. College of Electronic and Information Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, China; 2. Key Laboratory of Textile Science & Technology, Ministry of Education, Donghua University, Shanghai 201620, China; 3. Key La⁃ boratory of Advanced Textile Materials and Manufacturing Technology, Ministry of Education, Zhejiang Sci⁃Tech University, Hangzhou 310018, China) Abstract:Image noise reduction has been identified as the first step of automatic fabric defect detection. Its effect has a direct impact on subsequent image segmentation, feature extraction and the final recognition result. To im⁃ prove the performance of noise reduction, a method of noise reduction for a fabric defect image based on complex contourlet transform and anisotropic diffusion has been proposed. Firstly, a fabric defect image was decomposed into low⁃frequency and high⁃frequency components through complex contourlet transform. Next, a P _Laplace operator and Catte_PM model were used to diffuse low⁃frequency and high⁃frequency components, respectively. Finally, the defect image was reconstructed by inverse complex contourlet transform. A large number of experimental results in⁃ dicate that, compared with the hybrid method of wavelet threshold shrinkage with total variation diffusion, the meth⁃ od combining the wavelet with PM model diffusion, the method combining contourlet with total variation and adap⁃ tive contrast diffusion, and the method combining nonsubsampled contourlet with nonlinear diffusion, the proposed method has great improvement in both subjective visual effect and objective quantitative evaluation indicators, which can preserve the texture and details of fabric image better. The proposed method has stronger noise reduction capabilities and can suppress noise effectively. Keywords:fabric defect detection; fabric defect image; image noise reduction; complex contourlet transform; ani⁃ sotropic diffusion; P_Laplace operator; Catte_PM model 收稿日期:2012⁃11⁃08. 网络出版日期:2013⁃05⁃15. 基金项目:国家自然科学基金资助项目(60872065);纺织面料技术教 育部重点实验室开放基金资助项目( P1111);先进纺织材 料与制备技术教育部重点实验室开放课题研究基金资助 项目(2010001);江苏高校优势学科建设工程资助项目. 通信作者:吴一全. E⁃mail: nuaaimage@ 163.com. 织物疵点是纺织品生产过程中影响产品质量的 重要因素之一.传统的人工检测方法速度慢,劳动强 度大,易造成误检和漏检.基于图像处理技术的自动 检测系统可以极大地提高织物疵点的检测效率和精 度,减少人为因素制约,大大提高了纺织品生产流程 中的自动化程度.织物疵点图像获取和传输过程中 常受到高斯噪声和微量椒盐噪声的污染[1] ,致使图
向下翻页>>
©2008-现在 cucdc.com 高等教育资讯网 版权所有