Compressive Sensing is a new way of sampling signals at a sub-Nyquist rate. For many signals, this revolutionary technology strongly relies on the sparsity of the signal and incoherency between sensing basis and representation basis. In this work, compressed sensing method is proposed to reduce the noise of the image signal. Noise reduction and image reconstruction are formulated in the theoretical framework of compressed sensing using Basis Pursuit de-noising (BPDN) and Compressive Sampling Matching Pursuit (CoSaMP) algorithm when random measurement matrix is utilized to acquire the data. Ultimately, it is demonstrated that the proposed methods can't perfectly recover the image signal. Therefore, we have used a complementary approach for enhancing the performance of CS recovery with non-sparse signals. In this work, we have used a new designed CS recovery framework, called De-noising-based Approximate Message Passing (D-AMP). This method uses a de-noising algorithm to recover signals from compressive measurements. For de-noising purpose the Non-Local Means (NLM), Bayesian Least Squares Gaussian Scale Mixtures (BLS-GSM) and Block Matching 3D collaborative have been used. Also, in this work, we have evaluated the performance of our proposed image enhancement methods using the quality measure peak signal-to-noise ratio (PSNR).
%0 Journal Article
%1 ujan2017comparative
%A Ujan, Sahar
%A Ghorshi, Seyed
%A Pourebrahim, Majid
%D 2017
%J International Journal of Image Processing (IJIP)
%K (BL, (BM3D) (BP), (CoSaMP), (D-AMP), (NLM), 3D Approximate Basis Bayesian Block Compressive Gaussian Least Matching Means Message Mixtures Non-local Passing Pursuit Sampling Scale Sensing, Squares collaborative filter
%N 1
%P 106-120
%T Comparative Study of Compressive Sensing Techniques For Image Enhancement
%U http://www.cscjournals.org/library/manuscriptinfo.php?mc=IJIP-1129
%V 11
%X Compressive Sensing is a new way of sampling signals at a sub-Nyquist rate. For many signals, this revolutionary technology strongly relies on the sparsity of the signal and incoherency between sensing basis and representation basis. In this work, compressed sensing method is proposed to reduce the noise of the image signal. Noise reduction and image reconstruction are formulated in the theoretical framework of compressed sensing using Basis Pursuit de-noising (BPDN) and Compressive Sampling Matching Pursuit (CoSaMP) algorithm when random measurement matrix is utilized to acquire the data. Ultimately, it is demonstrated that the proposed methods can't perfectly recover the image signal. Therefore, we have used a complementary approach for enhancing the performance of CS recovery with non-sparse signals. In this work, we have used a new designed CS recovery framework, called De-noising-based Approximate Message Passing (D-AMP). This method uses a de-noising algorithm to recover signals from compressive measurements. For de-noising purpose the Non-Local Means (NLM), Bayesian Least Squares Gaussian Scale Mixtures (BLS-GSM) and Block Matching 3D collaborative have been used. Also, in this work, we have evaluated the performance of our proposed image enhancement methods using the quality measure peak signal-to-noise ratio (PSNR).
@article{ujan2017comparative,
abstract = {Compressive Sensing is a new way of sampling signals at a sub-Nyquist rate. For many signals, this revolutionary technology strongly relies on the sparsity of the signal and incoherency between sensing basis and representation basis. In this work, compressed sensing method is proposed to reduce the noise of the image signal. Noise reduction and image reconstruction are formulated in the theoretical framework of compressed sensing using Basis Pursuit de-noising (BPDN) and Compressive Sampling Matching Pursuit (CoSaMP) algorithm when random measurement matrix is utilized to acquire the data. Ultimately, it is demonstrated that the proposed methods can't perfectly recover the image signal. Therefore, we have used a complementary approach for enhancing the performance of CS recovery with non-sparse signals. In this work, we have used a new designed CS recovery framework, called De-noising-based Approximate Message Passing (D-AMP). This method uses a de-noising algorithm to recover signals from compressive measurements. For de-noising purpose the Non-Local Means (NLM), Bayesian Least Squares Gaussian Scale Mixtures (BLS-GSM) and Block Matching 3D collaborative have been used. Also, in this work, we have evaluated the performance of our proposed image enhancement methods using the quality measure peak signal-to-noise ratio (PSNR).},
added-at = {2018-12-14T11:05:45.000+0100},
author = {Ujan, Sahar and Ghorshi, Seyed and Pourebrahim, Majid},
biburl = {https://www.bibsonomy.org/bibtex/249846d8e137ed4a409f89ec8b7c033d2/cscjournals},
interhash = {3a5ce716bfba5dbe37e1832f620443cb},
intrahash = {49846d8e137ed4a409f89ec8b7c033d2},
issn = {1985-2304},
journal = {International Journal of Image Processing (IJIP)},
keywords = {(BL, (BM3D) (BP), (CoSaMP), (D-AMP), (NLM), 3D Approximate Basis Bayesian Block Compressive Gaussian Least Matching Means Message Mixtures Non-local Passing Pursuit Sampling Scale Sensing, Squares collaborative filter},
language = {English},
month = {August},
number = 1,
pages = {106-120},
timestamp = {2018-12-14T11:05:45.000+0100},
title = {Comparative Study of Compressive Sensing Techniques For Image Enhancement},
url = {http://www.cscjournals.org/library/manuscriptinfo.php?mc=IJIP-1129},
volume = 11,
year = 2017
}