基于曲波变换和色度模型的彩色图像去噪
|
摘要
针对彩色图像中高斯噪声难以去除的问题,提出了一种基于色度模型的色度中值滤波与曲波变换域Bayes多阈值滤波相结合的彩色图像去噪方法。该方法首先将RGB图像转换为HSI图像,通过在色度平面搜索色度中值分别消除H和S分量中的噪声,然后采用循环平移的曲波变换域Bayes多阈值对I分量噪声进行消除,最后通过HSI反变换得到去噪后的彩色图像。对比实验结果表明,本文方法对彩色图像高斯噪声去除,在获得较高峰值信噪比的同时,能较好保留彩色图像的色度、亮度和图像纹理细节信息,避免了"伪影"现象的产生。
Gaussian noise in color images is difficult to remove.We propose a denoising method by combining chrominance median filter based on chrominance model and Bayes multi-threshold filter in the curvelet transform domain.The method converts RGB images to HSI images first.The noise in the H and S components is filtered by searching chrominance median in the chrominance plane.Then,the noise in the I component is removed by using Bayes multi-threshold filter in the cycle spinning curvelet domain.Finally,the denoised color images are produced through the HSI inverse transformation.The contrast experiment results indicate that while achieving higher PSNR,the proposed method can better preserve the color images' chrominance,intensity as well as texture information,while avoiding yielding artifact effect for Gaussian noise removal in color images.
Gaussian noise in color images is difficult to remove.We propose a denoising method by combining chrominance median filter based on chrominance model and Bayes multi-threshold filter in the curvelet transform domain.The method converts RGB images to HSI images first.The noise in the H and S components is filtered by searching chrominance median in the chrominance plane.Then,the noise in the I component is removed by using Bayes multi-threshold filter in the cycle spinning curvelet domain.Finally,the denoised color images are produced through the HSI inverse transformation.The contrast experiment results indicate that while achieving higher PSNR,the proposed method can better preserve the color images' chrominance,intensity as well as texture information,while avoiding yielding artifact effect for Gaussian noise removal in color images.
引文
[1]JinL H,Yao X Z,Li D H.A survey on color image vector filte-ring techniques[J].Journal of Image and Graphics,2009,14(2):243-254.[金良海,姚行中,李德华.彩色图像矢量滤波技术综述[J].中国图象图形学报,2009,14(2):243-254.]
[2]Zeng G P.Research on image denoising based on Wavelet trans-form[D].Beijing:Beijing Jiaotong University,2007.[曾国平.基于小波变换的图像去噪算法研究[D].北京:北京交通大学,2007.]
[3]Andreas K,Mongi A.Digital Color Image Processing[M].Bei-jing:Tsinghua Press,2010:112-122.
[4]Candes E J,Donoho D L.Curvelet—A surprisingly effective non-adaptive representation for objects with edges[C]//Rabut C,Cohen A,Schumaker L L.Curves and Surfaces.Nashville,USA:Vanderbilt University Press,1999:105-120.
[5]Candes E J,Donoho D L.Recovering edges in ill-posed inverseproblems:Optimality of curvelet frames[J].The Annals of Statis-tics,2002,30(3):784-842.
[6]Candes E J,Demanet L,Donoho D L.Fast discrete curvelettransforms[R].California:California Institute of Technology,2005.
[7]Wang W B,Yi X M,Fei P S.An image denoising method basedon the relativity of the curvelet coefficients[J].Journal of Optoe-lectronics.Laser,2006,17(12):1519-1523.[王文波,羿旭明,费浦生.基于曲波系数相关性的去噪算法[J].光电子.激光,2006,17(12):1519-1523.]
[8]Wu A D,Zhao X L.Seismic signal denoising method based oncurvelet transform[J].Journal of China University of Petroleum,2010,34(3):30-33.[吴爱弟,赵秀玲.基于曲波变换的地震信号去噪方法[J].中国石油大学学报,2010,34(3):30-33.]
[9]Wang X L,Li L Y.Image denoising algorithm based on local sta-tistical property of curvelet coefficients[J].Journal of WuhanUniversity of Technology,2009,31(2):1-5.[汪祥莉,李腊元.基于曲波系数局部统计特性的图像去噪算法[J].武汉理工大学学报,2009,31(2):1-5.]
[10]He J,Li H W,Zhang Q.Curvelet transform algorithm with auto-adaptive threshold to image denoising[J].Journal of Data Acqui-sition&Processing,2010,25(3):336-340.[何劲,李宏伟,张群.一种自适应阈值曲波图像去噪算法[J].数据采集与处理,2010,25(3):336-340.]
[11]Zhang X C,Zhang C J.Satellite cloud image de-noising and en-hancement by fuzzy wavelet neural network and genetic algorithmin curvelet domain[J].Lecture Notes in Computer Science,2007,4688(1):389-395.
[12]Li J,Meng Z D.HS joint statistics based adaptive color segmenta-tion[J].Joural of Southeast University(Natural Science Edi-tion),2010,40(sup1):266-271.[李峻,孟正大.基于HS分量联合统计的自适应阈值分割算法[J].东南大学学报:自热科学版,2010,40(增刊1):266-271.]
[13]Nguyen T B,Ashish K.Multilevel threshold based image denois-ing in curvelet domain[J].Journal of Computer Science andTechnology,2010,25(3):632-640.
[14]Yang J Y.A de-noising algorithm for color images based on thesecond generation bandelet transtorm[J].Computer Engineering&Science,2010,32(6):65-67.[杨居义.基于第二代bandelet变换的彩色图像去噪算法[J].计算机工程与科学,2010,32(6):65-67.]
[15]Astola J,Haavisto P,Neuvo Y.Vector median filters[J].Pro-ceedings of the IEEE,1990,78(4):678-689.
[2]Zeng G P.Research on image denoising based on Wavelet trans-form[D].Beijing:Beijing Jiaotong University,2007.[曾国平.基于小波变换的图像去噪算法研究[D].北京:北京交通大学,2007.]
[3]Andreas K,Mongi A.Digital Color Image Processing[M].Bei-jing:Tsinghua Press,2010:112-122.
[4]Candes E J,Donoho D L.Curvelet—A surprisingly effective non-adaptive representation for objects with edges[C]//Rabut C,Cohen A,Schumaker L L.Curves and Surfaces.Nashville,USA:Vanderbilt University Press,1999:105-120.
[5]Candes E J,Donoho D L.Recovering edges in ill-posed inverseproblems:Optimality of curvelet frames[J].The Annals of Statis-tics,2002,30(3):784-842.
[6]Candes E J,Demanet L,Donoho D L.Fast discrete curvelettransforms[R].California:California Institute of Technology,2005.
[7]Wang W B,Yi X M,Fei P S.An image denoising method basedon the relativity of the curvelet coefficients[J].Journal of Optoe-lectronics.Laser,2006,17(12):1519-1523.[王文波,羿旭明,费浦生.基于曲波系数相关性的去噪算法[J].光电子.激光,2006,17(12):1519-1523.]
[8]Wu A D,Zhao X L.Seismic signal denoising method based oncurvelet transform[J].Journal of China University of Petroleum,2010,34(3):30-33.[吴爱弟,赵秀玲.基于曲波变换的地震信号去噪方法[J].中国石油大学学报,2010,34(3):30-33.]
[9]Wang X L,Li L Y.Image denoising algorithm based on local sta-tistical property of curvelet coefficients[J].Journal of WuhanUniversity of Technology,2009,31(2):1-5.[汪祥莉,李腊元.基于曲波系数局部统计特性的图像去噪算法[J].武汉理工大学学报,2009,31(2):1-5.]
[10]He J,Li H W,Zhang Q.Curvelet transform algorithm with auto-adaptive threshold to image denoising[J].Journal of Data Acqui-sition&Processing,2010,25(3):336-340.[何劲,李宏伟,张群.一种自适应阈值曲波图像去噪算法[J].数据采集与处理,2010,25(3):336-340.]
[11]Zhang X C,Zhang C J.Satellite cloud image de-noising and en-hancement by fuzzy wavelet neural network and genetic algorithmin curvelet domain[J].Lecture Notes in Computer Science,2007,4688(1):389-395.
[12]Li J,Meng Z D.HS joint statistics based adaptive color segmenta-tion[J].Joural of Southeast University(Natural Science Edi-tion),2010,40(sup1):266-271.[李峻,孟正大.基于HS分量联合统计的自适应阈值分割算法[J].东南大学学报:自热科学版,2010,40(增刊1):266-271.]
[13]Nguyen T B,Ashish K.Multilevel threshold based image denois-ing in curvelet domain[J].Journal of Computer Science andTechnology,2010,25(3):632-640.
[14]Yang J Y.A de-noising algorithm for color images based on thesecond generation bandelet transtorm[J].Computer Engineering&Science,2010,32(6):65-67.[杨居义.基于第二代bandelet变换的彩色图像去噪算法[J].计算机工程与科学,2010,32(6):65-67.]
[15]Astola J,Haavisto P,Neuvo Y.Vector median filters[J].Pro-ceedings of the IEEE,1990,78(4):678-689.
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心