Singular Value Decomposition (svd)

In this paper, a new no-reference image quality assessment (NR-IQA) metric for grey images is proposed using LIVE II image database. The features used are extracted from three well-known NR-IQA objective metrics based on natural scene statistical attributes from three different domains. These metrics may contain redundant, noisy or less informative features which affect the quality score prediction. In order to overcome this drawback, the first step of our work consists in selecting the most relevant image quality features by using Singular Value Decomposition (SVD) based dominant eigenvectors. The second step is performed by employing Relevance Vector Machine (RVM) to learn the mapping between the previously selected features and human opinion scores. Simulations demonstrate that the proposed metric performs very well in terms of correlation and monotonicity.

With the advancement in technology, digital data such as image, video, etc. can be easily manipulated. Image hashing is a method that can be used for authentica-tion and identification of digital images. In this communication , a robust image hashing technique is proposed using scale invariant feature transform (SIFT), singular value decomposition (SVD) and Zer-nike moment. Zernike moment is used to restore the image against the rotation attack. Potential points are generated from the image by using SIFT. Block processing of equal size is performed on the input image. Key points within the same block are used to generate hashing values. The experimental outcome shows that the proposed technique can withstand different types of attacks. Receiver operating characteristic curve comparison specifies that our method outper-forms other existing methods under consideration.

The purpose of this article is to determine the usefulness of the Graphics Processing Unit (GPU) calculations used to implement the Latent Semantic Indexing (LSI) reduction of the TERM-BY-DOCUMENT matrix. Considered reduction of the matrix is based on the use of the SVD (Singular Value Decomposition) decomposition. A high computational complexity of the SVD decomposition-O(n 3), causes that a reduction of a large indexing structure is a difficult task. In this article there is a comparison of the time complexity and accuracy of the algorithms implemented for two different environments. The first environment is associated with the CPU and MATLAB R2011a. The second environment is related to graphics processors and the CULA library. The calculations were carried out on generally available benchmark matrices, which were combined to achieve the resulting matrix of high size. For both considered environments computations were performed for double and single precision data.

Principal components analysis (PCA) is one of a family of techniques for taking high-dimensional data, and using the dependencies between the variables to represent it in a more tractable, lower-dimensional form, without losing too much information. PCA is one of the simplest and most robust ways of doing such dimensionality reduction. It is also one of the best, and has been rediscovered many times in many fields, so it is also known as the Karhunen-Lo_eve transformation, the Hotelling transformation, the method of empirical orthogonal functions, and singular value decomposition.

The objective of this paper is to compare the performance of Singular Value Decomposition (SVD) method and Radial Basis Function (RBF) Neural Network for optimization of fuzzy outputs in the epilepsy risk level classifications from EEG (Electroencephalogram) signals. The fuzzy pre classifier is used to classify the risk levels of epilepsy based on extracted parameters like energy, variance, peaks, sharp and spike waves, duration, events and covariance from the EEG signals of the patient. SVD and RBF neural network is exploited on the classified data to identify the optimized risk level (singleton) which characterizes the patient's epilepsy risk level. The efficacy of the above methods is compared based on the bench mark parameters such as Performance Index (PI) and Quality Value (QV).

This paper presents a hybrid watermarking technique for medical images. The method uses a combination of three transforms: Discrete Wavelet Transform (DWT), Discrete Cosine Transform (DCT), and singular value decomposition (SVD). Then, the paper discusses the results of applying the combined method on different medical images from eight patients. The images were watermarked with a small watermark image representing the patients' medical data. The visual quality of the watermarked images (before and after attacks) was analyzed using five quality metrics: PSNR, WSNR, PSNR-HVS-M, PSNR-HVS, and MSSIM. The first four metrics' average values of the watermarked medical images before attacks were approximately 32 db, 35 db, 42 db, and 40 db respectively; while the MSSM index indicated a similarity between the original and watermarked images of more than 97%. However, the metric values decreased significantly after attacking the images with various operations even though the watermark image could be retrieved after almost all attacks. In brief, the initial results indicate that watermarking medical images with patients' data does not significantly affect their visual quality and they can still be used by medical staff.

With the vast growth of internet technology, digital content protection has
become very important. Various algorithms have been/being developed for preserving
proprietary rights. Digital watermarking technique is the good measure for preserving
such rights. In this thesis, SVD based digital image watermarking using interpolation
method was done where the watermark was added in the singular values of a digital
image using interpolation method. The resultant watermarked image was treated with
different distortion operation. Finally the watermark was extracted for each method and
each recovered watermark was compared with the original watermark using normalized
correlation and accuracy rate. The better method was prescribed based on comparison
results. The selected method can be easily implemented in real world application for the
purpose of copy right protection, authentication, integrity etc
Keywords: Digital Watermarking, Copy right protection, Singular value decomposition

This paper focuses on no-reference image quality assessment(NR-IQA)metrics. In the literature, a wide range of algorithms are proposed to automatically estimate the perceived quality of visual data. However, most of them are not able to effectively quantify the various degradations and artifacts that the image may undergo. Thus, merging of diverse metrics operating in different information domains is hoped to yield better performances, which is the main theme of the proposed work. In particular, the metric proposed in this paper is based on three well-known NR-IQA objective metrics that depend on natural scene statistical attributes from three different domains to extract a vector of image features. Then, Singular Value Decomposition (SVD) based dominant eigenvectors method is used to select the most relevant image quality attributes. These latter are used as input to Relevance Vector Machine (RVM) to derive the overall quality index. Validation experiments are divided into two groups; in the first group, learning process (training and test phases) is applied on one single image quality database whereas in the second group of simulations, training and test phases are separated on two distinct datasets. Obtained results demonstrate that the proposed metric performs very well in terms of correlation, monotonicity and accuracy in both the two scenarios.

Secret key extraction is a crucial issue in physical layer security and a promising technique for the next generation of 5G and beyond. Unlike previous works on this topic in which Orthogonal Frequency Division Multiplexing (OFDM) sub-channels are considered to be independent, the effect of correlation among sub-channels on the secret key rate is addressed in this paper. We assume a realistic model for dependency among the sub-channels. Moreover, a new approach with low computational complexity for key extraction and optimal Mutual Information (MI) is presented in our study. To do this, we utilize a Singular Value Decomposition-based (SVD-based) precoding to obtain an optimal SVD-based approach for key extraction. Simulation results show that the rate of the key exchange link may drop up to 72% when the signal to noise ratio (SNR) is considered to be 8dB. The low computational complexity of our proposed approach makes it a promising candidate for developing secure and high-speed networks.

This paper is a comparative study between the performances of conventional terrestrial multicell Multiple Input Multiple Output (MIMO) working in a Rayleigh fading environment and its corresponding High Altitude Platform (HAP) system working under a different Path Loss (PL) model with the capacity as the performance metric of interest, calculated using the Singular Value Decomposition (SVD). Both systems are assumed to be affected by cochannel interference users in other cells. Simulation results show that the performance of multicell MIMO HAP dependent system outperforms its corresponding that works in terrestrial environment in terms of per user channel capacity.

Video denoising is a longstanding area of research. The core problem is to remove noise without affecting image details. The paper presents a new algorithm which combines motion compensation, patch based method and multiresolution bilateral filter for denoising. The proposed approach takes advantage of the self-similarity and redundancy of adjacent frames. Motion compensation by optical flow method permits the use of spatio-temporal patches for a more robust comparison. The use of singular value decomposition (SVD) for patches preserves fine details such as edges and texture. Multiresolution bilateral filter framework, which integrates both bilateral filter and soft thresholding, has the potential of eliminating low-frequency noise components. Experimental results shows that the proposed method eliminate most of the artifacts due to noise and improves texture and detail preservation.

We consider calibration of second-order, or hyphenated instruments generating bilinear two-way data for each specimen. The bilinear regression model is to be estimated from a number of specimens of known composition. We propose a simple estimator and study how it works on real and simulated data. The estimator, which we call the SVD singular value decom-.position estimator is usually not much less efficient than bilinear least squares. The advantages of our method over bilinear .least squares are that it is faster and more easily computed, its standard errors are explicit and derived in the paper, and it

Signal & Image Processing : An International Journal is an Open Access peer-reviewed journal intended for researchers from academia and industry, who are active in the multidisciplinary field of signal & image processing. The scope of the journal covers all theoretical and practical aspects of the Digital Signal Processing & Image processing, from basic research to development of application.

This paper explores and compares techniques for automatically recognizing facial actions in sequences of images.The comparative study of Facial Expression Recognition techniques namely Principal Component analysis (PCA), PCA with SVD (Singular Value Decomposition) is done.The objective of this research is to show that PCA with SVD is superior to former technique in terms of recognition rate.To test and evaluate their performance, experiments are performed using JAFEE and real database using both techniques. The universally accepted five principal emotions to be recognized are: Angry, Happy, Sad, Disgust and Surprise along with neutral. The recognition rate is obtained on all the facial expressions.

This paper presents a contourlet transform based watermarking scheme for colour images. It also makes use of discrete cosine transform and singular value decomposition. Discrete cosine transform is applied on three colour components (Y, Cb and Cr), obtained from the colour image, to produce DCT coefficients. These coefficients are decomposed into directional subbands using 2- level contourlet transform. The watermark is also decomposed using contourlet transform. The fourth directional subbands of cover image and watermark as well are further subjected to Laplacian pyramid decomposition. As a result, the low pass subbands are obtained. The singular values of lowpass subband of watermark are embedded into singular values of lowpass subband of cover image. The experimental results reveal that the proposed watermarking scheme is robust against common image processing attacks. The results also demonstrate good imperceptibility.

Today an increasing digital world, personal reliable authentication has become an important human Computer interface activity. It is very important to establish a person's identity. In today existing security mainly depends on passwords, swipe cards or token based approach and attitude to control access to physical and virtual spaces passport. Universal, such as methods, although very secure. Such as tokens, badges and access cards can be shared or stolen. Passwords and PIN numbers can be also stolen electronically. In addition, they cannot distinguish between authentic have access to or knowledge of the user and tokens. To make a system more secure and simple with the use of biometric authentication system such as face and hand gesture recognition for personal authentication. So in this paper, A Hidden Markov Model (HMM) based face recognition system using Singular Value Decomposition (SVD) is proposed. Neha Rana | Bhavna Pancholi"HMM-Based Face Recognition System with SVD...

With the advancement in technology, digital data such as image, video, etc. can be easily manipulated. Image hashing is a method that can be used for authentica-tion and identification of digital images. In this communication , a robust image hashing technique is proposed using scale invariant feature transform (SIFT), singular value decomposition (SVD) and Zer-nike moment. Zernike moment is used to restore the image against the rotation attack. Potential points are generated from the image by using SIFT. Block processing of equal size is performed on the input image. Key points within the same block are used to generate hashing values. The experimental outcome shows that the proposed technique can withstand different types of attacks. Receiver operating characteristic curve comparison specifies that our method outper-forms other existing methods under consideration.

This paper presents a novel hierarchical coding scheme for light fields based on transmittance patterns of low-rank multiplicative layers and Fourier disparity layers. The proposed scheme identifies multiplicative layers of light field view subsets optimized using a convolutional neural network for different scanning orders. Our approach exploits the hidden low-rank structure in the multiplicative layers obtained from the subsets of different scanning patterns. The spatial redundancies in the multiplicative layers can be efficiently removed by performing low-rank approximation at different ranks on the Krylov subspace. The intra-view and inter-view redundancies between approximated layers are further removed by HEVC encoding. Next, a Fourier disparity layer representation is constructed from the first subset of the approximated light field based on the chosen hierarchical order. Subsequent view subsets are synthesized by modeling the Fourier disparity layers that iteratively refine the representation with improved accuracy. The critical advantage of the proposed hybrid layered representation and coding scheme is that it utilizes not just spatial and temporal redundancies in light fields but efficiently exploits intrinsic similarities among neighboring sub-aperture images in both horizontal and vertical directions as specified by different predication orders. In addition, the scheme is flexible to realize a range of multiple bitrates at the decoder within a single integrated system. The compression performance of the proposed scheme is analyzed on real light fields. We achieved substantial bitrate savings and maintained good light field reconstruction quality.

Research in the fields of machine learning and intelligent systems addresses essential problem of developing computer algorithms that can deal with huge amounts of data and then utilize this data in an intellectual way to solve a variety of real-world problems. In many applications, to interpret data with a large number of variables in a meaningful way, it is essential to reduce the number of variables and interpret linear combinations of the data. Principal Component Analysis (PCA) is an unsupervised learning technique that uses sophisticated mathematical principles to reduce the dimensionality of large datasets. The goal of this paper is to provide a complete understanding of the sophisticated PCA in the fields of machine learning and data dimensional reduction. It explains its mathematical aspect and describes its relationship with Singular Value Decomposition (SVD) when PCA is calculated using the covariance matrix. In addition, with the use of MATLAB, the paper shows the usefulness of PCA in representing and visualizing Iris dataset using a smaller number of variables.

The purpose of this article is to determine the usefulness of the Graphics Processing Unit (GPU) calculations used to implement the Latent Semantic Indexing (LSI) reduction of the TERM-BY-DOCUMENT matrix. Considered reduction of the matrix is based on the use of the SVD (Singular Value Decomposition) decomposition. A high computational complexity of the SVD decomposition-O(n 3), causes that a reduction of a large indexing structure is a difficult task. In this article there is a comparison of the time complexity and accuracy of the algorithms implemented for two different environments. The first environment is associated with the CPU and MATLAB R2011a. The second environment is related to graphics processors and the CULA library. The calculations were carried out on generally available benchmark matrices, which were combined to achieve the resulting matrix of high size. For both considered environments computations were performed for double and single precision data.

The reactive power is control by using Matrix converter at variable speed permanent-magnet synchronous wind generator and improving the performance. The frequency and amplitude of induced voltage also changes due to variation of wind speed. A matrix convertor is used to overcome this problem. A generalized modulation technique concept related to singular value decomposition of the modulation matrix is used to model different modulation techniques and investigate their corresponding input reactive power capability. A new control method is proposed for the matrix converter based on this modulation technique, which uses active and reactive parts of the generator current to increase the control capability of the grid-side reactive current compared to conventional modulation methods. A new control structure is also proposed which can control the matrix converter and the grid-side maximum achievable reactive power for all wind speeds and power conditions generator reactive current is improved.