小波分析在医学超声图像去噪和增强中的应用
|
摘要
小波变换是近十几年发展起来的一种新的信号和图像处理工具。小波分析良好的时频特性决定了它在图像去噪和增强中具有广阔的应用前景，使得这一领域充满生机。
超声检查技术已成为医学临床诊断的重要手段之一。医学超声图像成像过程中产生的噪声降低了图像质量，影响了医生对疾病的诊断，故有必要抑制超声图像噪声和增强图像。
超声图像去噪和增强是超声图像处理的一个预处理过程，它是病变识别和分析的前提，在医学图像处理中，医学超声图像的去噪和增强的研究有着重要的意义。本文首先介绍了小波图像去噪和增强的现状，然后阐述了小波图像去噪和增强的理论基础，最后是利用小波变换的多分辨率特性，结合人眼的视觉特性，围绕小波图像去噪和增强的中心问题进行了研究，提出了相应的处理方法。本文主要内容有：
在医学超声图像噪声抑制方面：提出了基于贝叶斯估计的小波去噪方法和半软阈值小波图像去噪法。这两种方法，在图像的不同分辨率上，分别对小波系数进行不同的处理。半软阈值去噪法体现了将多分辨率分析和自适应处理有机结合的思想。实验结果表明本文的方法，在抑制噪声的同时尽可能多的保留对医生有用的图像边缘、细节信息，该去噪方法确实是行之有效的。
在超声图像增强方面，提出了先采用基于小波的高频增强法来增强图像细节再用非线性对比度增强的方法来改善图像视觉效果的增强方法，以及基于小波和模糊算法的图像增强方法。这些方法既增强了图像的细节特征又符合人眼的视觉特性，提高了图像的清晰度，有效地避免了平坦区域噪声的过增强问题。实验结果表明此方法具有一定的应用价值。
Wavelet transform is a nee signal and image processing tool developed in recent years. The wavelet analysis has excellent time and frequency feature, which hand a progmising application in image de-noising and enhancement, it make the field full of vitality force.
Ultrasonic detection technology has already been one of the important means of medical clinical diagnosis. Noise derived from the imaging degraded image quality and affected the detection rate of correctness. So noise must be removed from ultrasound image and enhance image should be enhanced.
Ultrasound image de-noising and enhancement are a pre-processing step in its processing, it is also the premise of disease recognition and analysis. Research on ultrasonic image de-noising and enhancement have important meaning. At first the present state of research on de-noising and enhancement based on wavelet are introduced in this paper, then we make a brief description of theoretical knowledge about image de-noising and enhancement via wavelet. Finally, both image de-noising and enhancement based on wavelet are mainly studied according to multi-resolution of wavelet analysis together with human vision, we proposed relevant methods. The contents are as follows:
⒈ In the aspect about noise removal of ultrasound image, we present wavelet de-noising methods based on Bayesian estimation and semi-soft threshold image de-noising. The two methods, we make different processing in diverse resolution of image. The method of semi-soft threshold embody the idea of multi-resolution together with adaptive process. Experiment result shows that the two methods can preserve some useful image edge and details to doctor as much as possible while removing noise, it also indicate that the two are simple and reliable.
⒉ In the aspect of ultrasound image enhancement. Two methods are presented. The first method is of high components strengthened based on wavelet to enhance image detail, then, image vision quality is improved through nonlinear contrast enhancement. The second one via wavelet together with fuzzy algorithm. The two methods not only enhance image details but also fit in with human vision, distinct of image is improved and noise over-enhancing can be voided in flatness area. The experiment results indicate that they have worthiness in practical application.
超声检查技术已成为医学临床诊断的重要手段之一。医学超声图像成像过程中产生的噪声降低了图像质量，影响了医生对疾病的诊断，故有必要抑制超声图像噪声和增强图像。
超声图像去噪和增强是超声图像处理的一个预处理过程，它是病变识别和分析的前提，在医学图像处理中，医学超声图像的去噪和增强的研究有着重要的意义。本文首先介绍了小波图像去噪和增强的现状，然后阐述了小波图像去噪和增强的理论基础，最后是利用小波变换的多分辨率特性，结合人眼的视觉特性，围绕小波图像去噪和增强的中心问题进行了研究，提出了相应的处理方法。本文主要内容有：
在医学超声图像噪声抑制方面：提出了基于贝叶斯估计的小波去噪方法和半软阈值小波图像去噪法。这两种方法，在图像的不同分辨率上，分别对小波系数进行不同的处理。半软阈值去噪法体现了将多分辨率分析和自适应处理有机结合的思想。实验结果表明本文的方法，在抑制噪声的同时尽可能多的保留对医生有用的图像边缘、细节信息，该去噪方法确实是行之有效的。
在超声图像增强方面，提出了先采用基于小波的高频增强法来增强图像细节再用非线性对比度增强的方法来改善图像视觉效果的增强方法，以及基于小波和模糊算法的图像增强方法。这些方法既增强了图像的细节特征又符合人眼的视觉特性，提高了图像的清晰度，有效地避免了平坦区域噪声的过增强问题。实验结果表明此方法具有一定的应用价值。
Wavelet transform is a nee signal and image processing tool developed in recent years. The wavelet analysis has excellent time and frequency feature, which hand a progmising application in image de-noising and enhancement, it make the field full of vitality force.
Ultrasonic detection technology has already been one of the important means of medical clinical diagnosis. Noise derived from the imaging degraded image quality and affected the detection rate of correctness. So noise must be removed from ultrasound image and enhance image should be enhanced.
Ultrasound image de-noising and enhancement are a pre-processing step in its processing, it is also the premise of disease recognition and analysis. Research on ultrasonic image de-noising and enhancement have important meaning. At first the present state of research on de-noising and enhancement based on wavelet are introduced in this paper, then we make a brief description of theoretical knowledge about image de-noising and enhancement via wavelet. Finally, both image de-noising and enhancement based on wavelet are mainly studied according to multi-resolution of wavelet analysis together with human vision, we proposed relevant methods. The contents are as follows:
⒈ In the aspect about noise removal of ultrasound image, we present wavelet de-noising methods based on Bayesian estimation and semi-soft threshold image de-noising. The two methods, we make different processing in diverse resolution of image. The method of semi-soft threshold embody the idea of multi-resolution together with adaptive process. Experiment result shows that the two methods can preserve some useful image edge and details to doctor as much as possible while removing noise, it also indicate that the two are simple and reliable.
⒉ In the aspect of ultrasound image enhancement. Two methods are presented. The first method is of high components strengthened based on wavelet to enhance image detail, then, image vision quality is improved through nonlinear contrast enhancement. The second one via wavelet together with fuzzy algorithm. The two methods not only enhance image details but also fit in with human vision, distinct of image is improved and noise over-enhancing can be voided in flatness area. The experiment results indicate that they have worthiness in practical application.
引文
[1] [日]谷口庆治 数字图像处理·应用篇[M]. 科学出版社,2002,7
[2] Alin Achim, Anastasios Bezerianos, Panagiotis Tsakalides. Novel Bayesian Multiscale method for speckle removal in medical ultrasound image. IEEE Trans.on medical image. 2001, 20(8):772-783
[3] Xuli Zong, Andrew F. Laine, Edward A. Geiser. Speckle reduction and contrast enhancement of echocardiograms via multiscale nonlinear processing. IEEE Trans.on medical image. 1998, 17(4):532-540
[4] 陆系群 陈 纯. 图像处理原理、技术与算法[M]. 浙江大学出版社,2001,8
[5] 王修信 胡维平 梁冬冬等. 小波变换在超声图像降噪处理中的应用[J].广西物理. 2002,(1)
[6]H. Guo, J. E. Odegard, M. Lang, R. A. Gopinath, I. W. Selesnick, and C. S. Burrus, “Waveletbased speckle reduction with application to SAR based ATD/R,” First Int’l Conf. on Image
Processing, vol. 1, pp. 75–79, Nov. 1994.
[7] Xuli Zong, Andrew F.Kaine & Edward A.Geiser. Speckle reduction and contrast enhancement of echocardiograms via multiscale nonlinear processing[J]. IEEEtrans. Med. Image.,.1998, 17(4):532-540
[8] David L. Donoho. Denoosing by soft thresholding [J]. IEEETrans. on information theory. 1995,5,41(3):613-627
[9] Mallat S., Zhong S. Characterization of signals from multiscale edges . IEEE Trans. on PAMI, 1992, 7, 14(7): 710-732
[10] Hua Xie, Leland E. Pierce, Fawwax T.Ulaby. SAR speckle reduction using wavelet denoising and Markov Random Field modeling. IEEE Trans on Geoscience and Romote Sensing . 2002, 40(10):2196-2203
[12] 黄晓凌 廖孟扬 覃家美等. 基于小波分析的X射线照片增强研究[J]. 武汉大学学报(自然科学版),1998,44(1):121-124
[13] 杨词银 尚海波 贾晨光等. 基于区域分割的自适应反锐化掩模算法[J]. 光学 精密工程, 2003,11(2):188-192
[14]卢丽敏 周海银. 一种基于遗传算法的图像增强方法[J]. 数学理论与应用, 2003,3,23(1):82-88
[15] 周德龙 赵志国 潘 泉等. 基于模糊集的图像增强算法研究[J]. 电子与信息学报, 2002,21(7):905-908
[16] 吴颖谦 施鹏飞. 基于小波变换的低对比度图像增强[J]. 红外与激光工程, 2003, 32(1):4-7
[17] Laine A.Mammo graphic feature enhancement by multiscale analuysis. IEEE Trans Medical Imaging, 1994,13(4):725～740.
[18]宫武鹏 王永仲 一种基于小波变换的红外图像对比度增强技术 国防科技大学学报 2000,6,22(6):117-119
[19]吴颖谦 方 涛 李聪亮 施鹏飞. 一种基于小波分析和人眼视觉特性的图像增强方法[J]. 数据采集与处理, 2003,3 ,18(1):17-21
[20]胡昌华 张军波 夏军等. 基于MATLAB的系统分析与设计[M]. 西安电子科技大学出版社, 2000.
[21] 勒中鑫. 数字图像信息处理[M]. 国防工业出版社, 2003,1,第1版
[22] 胡广书. 数字信号处理——理论、算法与实现[M]. 清华大学出版社. 1997,8
[23] 程佩青. 数字信号处理教程[M]. 清华大学出版社，2001，8 p89
[24] 张兆礼 赵春晖 梅晓丹[M]. 现代图像处理技术及Matlab实现[M], 2001,11
[25] 程正兴. 小波分析算法与应用[M]. 西安交通大学出版社, 2001.p9
[27]朱秀昌 刘 峰 胡 栋. 数字图像处理与图像通信[M]. 北京邮电大学出版社，2002，5
[29] Kenneth.R.Castleman 著 朱志刚 等译. 数字图像处理[M]. 电子工业出版社,1998,2
[31] 王建中 张 晖 吴 斌 程文华. 基于Daubechies小波和中值滤波的图像去噪法[J]. 武汉理工大学学 报2001,3, 23(3): 19-25
[32] X.Huang,G.A.Woolsey. Image denoising using Wiener filtering and wavelet thresholding. IEEE International Conference on,2000,3: 1759-1762
[33] Haiguang Chen, Andrew Li, Leon Kaufman and James Hale. A fast filtering algorithm for image enhancement. IEEE Trans on medical imaging. 1994,9, 13(3):557-564
[34] Xuli Zong, Edward A.Geiser, Andrew F.Laine, David C.Wilson. Homomorphic wavelet shrinkage and feature emphasis for speckle reductiong and enhancement of echocardiographic images[J]. Image Processing, Proceedings of SPIE, 1996, 2710(12):658-667.
[35] Khaled Z. Abd-Elmoniem, Abou-Bakr M. Youssef, and Yasser M. Kadah Real-Time Speckle Reduction and Coherence Enhancement in Ultrasound Imaging via NonlinearAnisotropic Diffusion. IEEE Trans. On Biomedical Engineering, 2002,9, 49(9):997-1014.
[36] 孙兆林. MATLAB 6. x图像处理[M]. 清华大学出版社. 2002,5
[37]张伟. 基于高阶统计量的小波阈值去噪方法的研究[硕士]. 吉林大学, 2000,10.
[38] Maarten Jansen, Maurits Malfait, Adhermar Bultheel. Generalized cross validation for wavelet thresholding[J]. Signal processing , 1997,56: 33-44
[39]谢杰成 张大力 徐文立. 小波图像去噪综述[J]. 中国图象图形学报, 2002,7(A版)(3):209-217
[40] Jansen M, Malfait M, Bultheel A. Generalized cross validation for wavelet thresholding [J]. Signal Processing ,1997,56(1):33-44.
[43]Mario A. T. Figueiredo, Robert D. Nowak. Wavelet-based image estimation: an empirical Bayes approach using Jeffreys’ noninformative prior[J]. IEEE Trans. On image processing , 2001,9,10(9):1322-1331
[44] 边肇祺 张学工. 模式识别[M]. 清华大学出版社, 2000, 1第一版
[45] 胡战虎. 基于贝叶斯估计的多分辨图像滤波方法[J]. 电子学报, 2002,1,30(1): 66-68
[46]谢杰成 张大力 徐文立. 一种小波去噪方法的几点改进[J]. 清华大学学报(自然科学版). 2002,9, 42(9):1269-1272
[47]J.Bernardo and A.Smith. Beyesian theory. Chichester, U.K.: Wiley, 1994
[48] 张新明 沈兰荪. 基于小波和统计特性的自适应图像增强[J]. 信号处理， 2001, 6, 17(3): 227-231
[49] 张新明 沈兰荪. 基于小波的同态滤波器用于图像对比度增强[J].电子学报, 2001,29(4):531-533
[50] 石江宏. 基于模糊算法的图象处理新方法[J]. 福州大学学报(自然科学版), 1998, 26(4):43-47
[51] 汪亚明 丁 益 刘 峰. 基于一种新型模糊增强算法的眼底图像增强[J]. 北京生物医 学工程,18(3):129-133
[52] S. K. Pal and R. A. King, Image Enhancement Using Smoothing with Fuzzy Sets[J]. IEEE Trans. Syst., Man, Cybern., SMC-11, 1981,11(7):494-501.
[53] 吴国雄 陈武凡. 图像的模糊增强与聚类分析[J]. 小型微型计算机系统, 1994,11，15(11):21-26
[54] 雷向康. 一种改进的图像模糊增强方法[J]. 系统工程与电子技术,1997(12):21-23
[55]蔡汉添. 小波变换域中图像噪声平滑技术[J]. 光学技术. 1998,(6):6-9
[57] 谢杰成 张大力 徐文立. 一种小波去噪方法的几点改进[J]. 清华大学学报(自然科学版). 2002,9, 42(9):1269-1272
[58] 许雷 郑筱祥. 一种基于小波变换及数学形态学方法的眼底图像增强及定量分析方法[J]. 生物物理学报, 1998,3, 14(1): 70-76
[59]王修信 梁冬冬 胡维平等. 医学数字图像增强方法的研究[J]. 广西师范大学学报(自然科学版), 2002，20(3):23-26
[60] 周鸣争 李 华 汪 军. 基于软阈值小波痕迹图像去噪算法及应用[J]. 安徽工程科技学院学报,2003,12, 18(4):30-32
[61] 顾海军 林明秀 宋建中等. 图像采集卡硬件参数基于熵的自动优化[J]. 光学精密工程, 2003,10, 11(5):523-526
[2] Alin Achim, Anastasios Bezerianos, Panagiotis Tsakalides. Novel Bayesian Multiscale method for speckle removal in medical ultrasound image. IEEE Trans.on medical image. 2001, 20(8):772-783
[3] Xuli Zong, Andrew F. Laine, Edward A. Geiser. Speckle reduction and contrast enhancement of echocardiograms via multiscale nonlinear processing. IEEE Trans.on medical image. 1998, 17(4):532-540
[4] 陆系群 陈 纯. 图像处理原理、技术与算法[M]. 浙江大学出版社,2001,8
[5] 王修信 胡维平 梁冬冬等. 小波变换在超声图像降噪处理中的应用[J].广西物理. 2002,(1)
[6]H. Guo, J. E. Odegard, M. Lang, R. A. Gopinath, I. W. Selesnick, and C. S. Burrus, “Waveletbased speckle reduction with application to SAR based ATD/R,” First Int’l Conf. on Image
Processing, vol. 1, pp. 75–79, Nov. 1994.
[7] Xuli Zong, Andrew F.Kaine & Edward A.Geiser. Speckle reduction and contrast enhancement of echocardiograms via multiscale nonlinear processing[J]. IEEEtrans. Med. Image.,.1998, 17(4):532-540
[8] David L. Donoho. Denoosing by soft thresholding [J]. IEEETrans. on information theory. 1995,5,41(3):613-627
[9] Mallat S., Zhong S. Characterization of signals from multiscale edges . IEEE Trans. on PAMI, 1992, 7, 14(7): 710-732
[10] Hua Xie, Leland E. Pierce, Fawwax T.Ulaby. SAR speckle reduction using wavelet denoising and Markov Random Field modeling. IEEE Trans on Geoscience and Romote Sensing . 2002, 40(10):2196-2203
[12] 黄晓凌 廖孟扬 覃家美等. 基于小波分析的X射线照片增强研究[J]. 武汉大学学报(自然科学版),1998,44(1):121-124
[13] 杨词银 尚海波 贾晨光等. 基于区域分割的自适应反锐化掩模算法[J]. 光学 精密工程, 2003,11(2):188-192
[14]卢丽敏 周海银. 一种基于遗传算法的图像增强方法[J]. 数学理论与应用, 2003,3,23(1):82-88
[15] 周德龙 赵志国 潘 泉等. 基于模糊集的图像增强算法研究[J]. 电子与信息学报, 2002,21(7):905-908
[16] 吴颖谦 施鹏飞. 基于小波变换的低对比度图像增强[J]. 红外与激光工程, 2003, 32(1):4-7
[17] Laine A.Mammo graphic feature enhancement by multiscale analuysis. IEEE Trans Medical Imaging, 1994,13(4):725～740.
[18]宫武鹏 王永仲 一种基于小波变换的红外图像对比度增强技术 国防科技大学学报 2000,6,22(6):117-119
[19]吴颖谦 方 涛 李聪亮 施鹏飞. 一种基于小波分析和人眼视觉特性的图像增强方法[J]. 数据采集与处理, 2003,3 ,18(1):17-21
[20]胡昌华 张军波 夏军等. 基于MATLAB的系统分析与设计[M]. 西安电子科技大学出版社, 2000.
[21] 勒中鑫. 数字图像信息处理[M]. 国防工业出版社, 2003,1,第1版
[22] 胡广书. 数字信号处理——理论、算法与实现[M]. 清华大学出版社. 1997,8
[23] 程佩青. 数字信号处理教程[M]. 清华大学出版社，2001，8 p89
[24] 张兆礼 赵春晖 梅晓丹[M]. 现代图像处理技术及Matlab实现[M], 2001,11
[25] 程正兴. 小波分析算法与应用[M]. 西安交通大学出版社, 2001.p9
[27]朱秀昌 刘 峰 胡 栋. 数字图像处理与图像通信[M]. 北京邮电大学出版社，2002，5
[29] Kenneth.R.Castleman 著 朱志刚 等译. 数字图像处理[M]. 电子工业出版社,1998,2
[31] 王建中 张 晖 吴 斌 程文华. 基于Daubechies小波和中值滤波的图像去噪法[J]. 武汉理工大学学 报2001,3, 23(3): 19-25
[32] X.Huang,G.A.Woolsey. Image denoising using Wiener filtering and wavelet thresholding. IEEE International Conference on,2000,3: 1759-1762
[33] Haiguang Chen, Andrew Li, Leon Kaufman and James Hale. A fast filtering algorithm for image enhancement. IEEE Trans on medical imaging. 1994,9, 13(3):557-564
[34] Xuli Zong, Edward A.Geiser, Andrew F.Laine, David C.Wilson. Homomorphic wavelet shrinkage and feature emphasis for speckle reductiong and enhancement of echocardiographic images[J]. Image Processing, Proceedings of SPIE, 1996, 2710(12):658-667.
[35] Khaled Z. Abd-Elmoniem, Abou-Bakr M. Youssef, and Yasser M. Kadah Real-Time Speckle Reduction and Coherence Enhancement in Ultrasound Imaging via NonlinearAnisotropic Diffusion. IEEE Trans. On Biomedical Engineering, 2002,9, 49(9):997-1014.
[36] 孙兆林. MATLAB 6. x图像处理[M]. 清华大学出版社. 2002,5
[37]张伟. 基于高阶统计量的小波阈值去噪方法的研究[硕士]. 吉林大学, 2000,10.
[38] Maarten Jansen, Maurits Malfait, Adhermar Bultheel. Generalized cross validation for wavelet thresholding[J]. Signal processing , 1997,56: 33-44
[39]谢杰成 张大力 徐文立. 小波图像去噪综述[J]. 中国图象图形学报, 2002,7(A版)(3):209-217
[40] Jansen M, Malfait M, Bultheel A. Generalized cross validation for wavelet thresholding [J]. Signal Processing ,1997,56(1):33-44.
[43]Mario A. T. Figueiredo, Robert D. Nowak. Wavelet-based image estimation: an empirical Bayes approach using Jeffreys’ noninformative prior[J]. IEEE Trans. On image processing , 2001,9,10(9):1322-1331
[44] 边肇祺 张学工. 模式识别[M]. 清华大学出版社, 2000, 1第一版
[45] 胡战虎. 基于贝叶斯估计的多分辨图像滤波方法[J]. 电子学报, 2002,1,30(1): 66-68
[46]谢杰成 张大力 徐文立. 一种小波去噪方法的几点改进[J]. 清华大学学报(自然科学版). 2002,9, 42(9):1269-1272
[47]J.Bernardo and A.Smith. Beyesian theory. Chichester, U.K.: Wiley, 1994
[48] 张新明 沈兰荪. 基于小波和统计特性的自适应图像增强[J]. 信号处理， 2001, 6, 17(3): 227-231
[49] 张新明 沈兰荪. 基于小波的同态滤波器用于图像对比度增强[J].电子学报, 2001,29(4):531-533
[50] 石江宏. 基于模糊算法的图象处理新方法[J]. 福州大学学报(自然科学版), 1998, 26(4):43-47
[51] 汪亚明 丁 益 刘 峰. 基于一种新型模糊增强算法的眼底图像增强[J]. 北京生物医 学工程,18(3):129-133
[52] S. K. Pal and R. A. King, Image Enhancement Using Smoothing with Fuzzy Sets[J]. IEEE Trans. Syst., Man, Cybern., SMC-11, 1981,11(7):494-501.
[53] 吴国雄 陈武凡. 图像的模糊增强与聚类分析[J]. 小型微型计算机系统, 1994,11，15(11):21-26
[54] 雷向康. 一种改进的图像模糊增强方法[J]. 系统工程与电子技术,1997(12):21-23
[55]蔡汉添. 小波变换域中图像噪声平滑技术[J]. 光学技术. 1998,(6):6-9
[57] 谢杰成 张大力 徐文立. 一种小波去噪方法的几点改进[J]. 清华大学学报(自然科学版). 2002,9, 42(9):1269-1272
[58] 许雷 郑筱祥. 一种基于小波变换及数学形态学方法的眼底图像增强及定量分析方法[J]. 生物物理学报, 1998,3, 14(1): 70-76
[59]王修信 梁冬冬 胡维平等. 医学数字图像增强方法的研究[J]. 广西师范大学学报(自然科学版), 2002，20(3):23-26
[60] 周鸣争 李 华 汪 军. 基于软阈值小波痕迹图像去噪算法及应用[J]. 安徽工程科技学院学报,2003,12, 18(4):30-32
[61] 顾海军 林明秀 宋建中等. 图像采集卡硬件参数基于熵的自动优化[J]. 光学精密工程, 2003,10, 11(5):523-526