基于改进二维Haar小波的图像去噪算法
|
摘要
小波分解层数和阈值函数的选择会影响图像去噪的性能,为此提出了一种改进二维Haar小波阈值法实现图像去噪。该方法使用子带标准差来确定二维Haar小波变换后高频子带中信号能量的强弱,并以此决定是否进行下一层的小波分解。提出一种新的阈值函数,该阈值函数是连续的,可以克服硬阈值函数对于小波系数过度收缩的缺点,以及软阈值处理使图像边缘模糊的缺点,能在噪声小波系数和噪声之间提供更平滑的过渡图像信号小波系数。实验结果表明:所提方法在峰值信噪比(PSNR)和均方误差(MSE)方面优于其他方法。
The selection of wavelet decomposition levels and threshold function affects the performance of image denoising. An improved 2D Haar wavelet threshold method was proposed to realize image denoising in this paper. This method used sub-bands standard deviation to determine the signal energy in high frequency sub-bands after 2D Haar wavelet transform,and then decided whether to perform the next level of wavelet decomposition or not. In addition,a new threshold function was proposed,which is continuous. It can overcome the shortcomings of the hard threshold function that shrinks the wavelet coefficients excessively,and the soft threshold processing that blurs the edge of the image. It can provide a smoother wavelet coefficients between noise and wavelet coefficients of the transitional image signal. The experimental results show that the proposed method is superior to other methods in in peak signal-to-noise ratio( PSNR) and mean square error( MSE).
The selection of wavelet decomposition levels and threshold function affects the performance of image denoising. An improved 2D Haar wavelet threshold method was proposed to realize image denoising in this paper. This method used sub-bands standard deviation to determine the signal energy in high frequency sub-bands after 2D Haar wavelet transform,and then decided whether to perform the next level of wavelet decomposition or not. In addition,a new threshold function was proposed,which is continuous. It can overcome the shortcomings of the hard threshold function that shrinks the wavelet coefficients excessively,and the soft threshold processing that blurs the edge of the image. It can provide a smoother wavelet coefficients between noise and wavelet coefficients of the transitional image signal. The experimental results show that the proposed method is superior to other methods in in peak signal-to-noise ratio( PSNR) and mean square error( MSE).
引文
[1]GLOVER G H,LI T Q,RESS D.Image-based method for retrospective correction of physiological motion effects in f MRI:RETROICOR[J].Magnetic Resonance in Medicine,2015,44(1):162-167.
[2]KARPATHY A,FEIFEI L.Deep Visual-Semantic Alignments for Generating Image Descriptions[J].IEEETransactions on Pattern Analysis&Machine Intelligence,2017,39(4):664-676.
[3]ARAKERI M P,KUMAR B P V,BARSAIYA S,et al.Computer vision based robotic weed control system for precision agriculture[C]//Advances in Computing,Communications and Informatics(ICACCI),2017 International Conference on.Udupi,India:IEEE,2017:1201-1205.
[4]CHEN B H,HUANG S C.An advanced moving object detection algorithm for automatic traffic monitoring in real-world limited bandwidth networks[J].IEEE Transactions on Multimedia,2014,16(3):837-847.
[5]JOSHI S,KULKARNI R K.Diffusion based multiresolution filtering algorithms for accurate abnormility detection in medical images[C]//Advances in Computing,Communication and Control(ICAC3),2017 International Conference on.Mumbai,India:IEEE,2017:1-5.
[6]ZHANG G,SUI H,SONG Z,et al.Automatic Registration Method of SAR and Optical Image Based on Line Features and Spectral Graph Theory[C]//Multimedia and Image Processing(ICMIP),2017 2nd International Conference on.Wuhan,China:IEEE,2017:64-67.
[7]XIE X,JONES M,TAM G.Recognition,Tracking,and Optimisation[J].International Journal of Computer Vision,2017,122(3):409-410.
[8]KIM S H,CHOI H L.Moving target tracking and recognition method for unmanned airborne surveillance systems[J].Journal of Institute of Control,Robotics and Systems,2017,23(3):157-164.
[9]蒋辉,汪辉,张家树.梯度双边滤波的图像去噪[J].计算机工程与应用,2016,52(5):231-235.
[10]NAIMI H,ADAMOU-MITICHE A B H,Mitiche L.Medical image denoising using dual tree complex thresholding wavelet transform and Wiener filter[J].Journal of King Saud University-Computer and Information Sciences,2015,27(1):40-45.
[11]LUO E,CHAN S H,NGUYEN T Q.Adaptive image denoising by targeted databases[J].IEEE Transactions on Image Processing,2015,24(7):2167-2181.
[12]CHEN R,LIU X,ZHAI D,et al.Depth Image Denoising via Collaborative Graph Fourier Transform[C]//International Forum on Digital TV and Wireless Multimedia Communications.Singapore:Springer,2017:128-137.
[13]BENABDELKADER S,SOLTANI O.Wavelet image denoising based spatial noise estimation[C]//Signal Processing and Intelligent Systems Conference(SPIS),2015.Tehran,Iran:IEEE,2015:83-87.
[14]ZHANG F,LIU Z.Image Denoising Based on the Bivariate Model of Dual Tree Complex Wavelet Transform[C]//Computational Intelligence and Security(CIS),2015 11th International Conference on.Shenzhen,China:IEEE,2015:171-174.
[15]ISMAEL S H,MUSTAFA F M,OKüMüS I T.A New Approach of Image Denoising Based on Discrete Wavelet Transform[C]//Computer Applications&Research(WSCAR),2016 World Symposium on.USA:IEEE,2016:36-40.
[16]SRIVASTAVA M,ANDERSON C L,FREED J H.A new wavelet denoising method for selecting decomposition levels and noise thresholds[J].IEEE Access,2016(4):3862-3877.
[2]KARPATHY A,FEIFEI L.Deep Visual-Semantic Alignments for Generating Image Descriptions[J].IEEETransactions on Pattern Analysis&Machine Intelligence,2017,39(4):664-676.
[3]ARAKERI M P,KUMAR B P V,BARSAIYA S,et al.Computer vision based robotic weed control system for precision agriculture[C]//Advances in Computing,Communications and Informatics(ICACCI),2017 International Conference on.Udupi,India:IEEE,2017:1201-1205.
[4]CHEN B H,HUANG S C.An advanced moving object detection algorithm for automatic traffic monitoring in real-world limited bandwidth networks[J].IEEE Transactions on Multimedia,2014,16(3):837-847.
[5]JOSHI S,KULKARNI R K.Diffusion based multiresolution filtering algorithms for accurate abnormility detection in medical images[C]//Advances in Computing,Communication and Control(ICAC3),2017 International Conference on.Mumbai,India:IEEE,2017:1-5.
[6]ZHANG G,SUI H,SONG Z,et al.Automatic Registration Method of SAR and Optical Image Based on Line Features and Spectral Graph Theory[C]//Multimedia and Image Processing(ICMIP),2017 2nd International Conference on.Wuhan,China:IEEE,2017:64-67.
[7]XIE X,JONES M,TAM G.Recognition,Tracking,and Optimisation[J].International Journal of Computer Vision,2017,122(3):409-410.
[8]KIM S H,CHOI H L.Moving target tracking and recognition method for unmanned airborne surveillance systems[J].Journal of Institute of Control,Robotics and Systems,2017,23(3):157-164.
[9]蒋辉,汪辉,张家树.梯度双边滤波的图像去噪[J].计算机工程与应用,2016,52(5):231-235.
[10]NAIMI H,ADAMOU-MITICHE A B H,Mitiche L.Medical image denoising using dual tree complex thresholding wavelet transform and Wiener filter[J].Journal of King Saud University-Computer and Information Sciences,2015,27(1):40-45.
[11]LUO E,CHAN S H,NGUYEN T Q.Adaptive image denoising by targeted databases[J].IEEE Transactions on Image Processing,2015,24(7):2167-2181.
[12]CHEN R,LIU X,ZHAI D,et al.Depth Image Denoising via Collaborative Graph Fourier Transform[C]//International Forum on Digital TV and Wireless Multimedia Communications.Singapore:Springer,2017:128-137.
[13]BENABDELKADER S,SOLTANI O.Wavelet image denoising based spatial noise estimation[C]//Signal Processing and Intelligent Systems Conference(SPIS),2015.Tehran,Iran:IEEE,2015:83-87.
[14]ZHANG F,LIU Z.Image Denoising Based on the Bivariate Model of Dual Tree Complex Wavelet Transform[C]//Computational Intelligence and Security(CIS),2015 11th International Conference on.Shenzhen,China:IEEE,2015:171-174.
[15]ISMAEL S H,MUSTAFA F M,OKüMüS I T.A New Approach of Image Denoising Based on Discrete Wavelet Transform[C]//Computer Applications&Research(WSCAR),2016 World Symposium on.USA:IEEE,2016:36-40.
[16]SRIVASTAVA M,ANDERSON C L,FREED J H.A new wavelet denoising method for selecting decomposition levels and noise thresholds[J].IEEE Access,2016(4):3862-3877.