Singular Value Decomposition Features Research Articles

Perception-guided defocus blur detection based on SVD feature

Defocus blur detection aims to identify out-of-focus regions in a single image. Although defocus blur detection has attracted more and more attention in recent years, it still faces some challenges. In particular, the in-focus regions with low contrast are easily misidentified as out-of-focus regions. To address this problem, a perception-guided defocus blur detection method is proposed to estimate defocus blur amounts at edge locations from a single image based on the singular value decomposition (SVD) features. Inspired by the fact that blurring the clear region produces significant changes in SVD domain as compared with re-blurring the blurred one, new SVD features are extracted based on re-blurred singular value difference (RESVD) of the corresponding local gradient patches. Then, perceptual weight based on Just Noticeable Blur (JNB) is introduced to guide the sparse blur map estimation obtained with the SVD features. Finally, the full defocus blur map is constructed from the sparse defocus blur map by the Matting Laplace algorithm. Visual evaluations are conducted on the CUHK, DUT and CTCUG datasets, employing mean absolute error (MAE) and Fβ-measure as quantitative evaluation metrics. The experimental results demonstrate that the proposed defocus blur detection method is superior to 13 state-of-the-art methods. On the DUT dataset, the proposed method yields a high Fβ-measure (0.802) with a low MAE (0.081) compared to other methods (Fβ≤0.799, MAE ≥0.099). On the CUHK and CTCUG datasets, our method achieves the best balance between Fβ-measure and MAE. Furthermore, our method results in better visual effects than other methods regarding realism and quality.

Recommendation for Classification of News Categories Using Support Vector Machine Algorithm with SVD

Online news is a digital information media currently has a very easy and flexible updating process. The News Document grouping process is implemented in several stages, including Text Mining which includes Text Pre-processing which includes Tokenizing, Stopword removal, Stemming, Word Merging, TF-IDF and Confusion Matrix. Of the several techniques in Text Mining, the most frequently used for News Document classification is the Support Vector Machine (SVM). SVM has the advantage of being able to identify separate hyperplane that maximizes the margin between two or more different classes. The selection of features in SVM significantly affects the classification accuracy results. Therefore, in this study a combination of feature selection methods is used, namely Singular Value Decomposition in order to increase accuracy and reduce the Classifier Time Support Vector Machine. This research resulted in text classification in the form of categories Entertainment, Health, Politics and Technology. Based on the Support Vector Machines Algorithm, an accuracy rate of 81% was obtained with 360 Data Training and 120 Data Testing, after adding the Singular Value Decomposition feature with a K- Rank value of 50%, a significant increase in accuracy was obtained with an accuracy value of 94% and The time of Algorithm process is faster.

Modeling and analysis of tool wear prediction based on SVD and BiLSTM

Wear is one of the main forms of tool failure during machining. The prediction of tool wear is of great significance for ensuring the high quality of the workpiece. In order to improve prediction accuracy of tool wear, a tool wear prediction model based on singular value decomposition (SVD) and bidirectional long short-term memory neural network (BiLSTM) is proposed. The cutting force signal is taken as the monitoring signal. Firstly, the raw cutting force signal is reconstructed by Hankle matrix, and the SVD of the reconstructed matrix is performed to extract the signal features. Then, SVD features of the current sampling period and the previous four sampling periods are taken as the input, and the tool wear prediction value at the current time is obtained based on the BiLSTM. The experimental results show that the proposed SVD-BiLSTM model can effectively predict the tool wear and obtain higher prediction accuracy than other comparison models.

Computerized approach for cardiovascular risk level detection using photoplethysmography signals

Cardiovascular disease (CVD) is the leading cause of death globally. In order to decrease the medical cost for treating the heart related pathologies, this paper proposes a computer-aided diagnostic system to classify various risk level of cardiovascular disease utilizing inexpensive and non-obtrusive diagnostic tool called photoplethysmography (PPG). In this study, features such as singular value decomposition (SVD), statistical features and wavelets (Haar, Daubechies, and Symlet) are extracted from the photoplethysmography signals. These feature vectors are then applied to the softmax discriminant classifier (SDC) and Gaussian mixture model classifier (GMM) for classification of various risk phases of CVDs. The classification performance of the proposed model incorporating SDC with SVD and statistical feature vectors increases with sensitivity of 97.24%, specificity of 99.09% and an accuracy of 97.88%.The method presented in this paper assist cardiologists to validate their diagnosis.

Fundus Image Classification Using VGG-19 Architecture with PCA and SVD

Automated medical image analysis is an emerging field of research that identifies the disease with the help of imaging technology. Diabetic retinopathy (DR) is a retinal disease that is diagnosed in diabetic patients. Deep neural network (DNN) is widely used to classify diabetic retinopathy from fundus images collected from suspected persons. The proposed DR classification system achieves a symmetrically optimized solution through the combination of a Gaussian mixture model (GMM), visual geometry group network (VGGNet), singular value decomposition (SVD) and principle component analysis (PCA), and softmax, for region segmentation, high dimensional feature extraction, feature selection and fundus image classification, respectively. The experiments were performed using a standard KAGGLE dataset containing 35,126 images. The proposed VGG-19 DNN based DR model outperformed the AlexNet and spatial invariant feature transform (SIFT) in terms of classification accuracy and computational time. Utilization of PCA and SVD feature selection with fully connected (FC) layers demonstrated the classification accuracies of 92.21%, 98.34%, 97.96%, and 98.13% for FC7-PCA, FC7-SVD, FC8-PCA, and FC8-SVD, respectively.

Automated Tumor Segmentation Based on Hidden Markov Classifier using Singular Value Decomposition Feature Extraction in Brain MR images

ntroduction: Diagnosing brain tumor is not always easy for doctors, and existence of an assistant that facilitates the interpretation process is an asset in the clinic. Computer vision techniques are devised to aid the clinic in detecting tumors based on a database of tumor contained images. These precise techniques open up the way for recognizing a tumor even in the early stages of cancer development depicted as very low contrast object in an image. To this end, at first a set of similarity measures are calculated for each pixel in the image database. Then, this dataset is fed into a classifier that defines the mathematical rules necessary for making decision about a new image. After defining the rules, a model is formed which is ready to make decisions for new images outside the database. This decision is accompanied with an accuracy which is calculated in the end as the classifier performance measures. Materials and Methods: In this study, Singular Value Decomposition – SVD – method is used as a tool to extract similarity measures for each malignant image in the database. These measures form a matrix with which the classifier is trained. The classifier in this study, is a model based on Hidden Markov Random Field – HMRF – idea, that utilizes Bayes’ rule and maximum a posteriori criterion to obtain labels for each pixel and making decisions about he new images respectively. After sorting the new image as either malignant or not, the malignant image goes through a wavelet process for tumor segmentation. In this section, the image is transformed into a multilevel wavelet structure. Image segmentation is done using direct and inverse wavelet transform. Results: The classification performance is done quantitatively and qualitatively by calculating Volume Overlap Ratio (VOR), Recognition Rate (RR), True Negative Rate (TNR), True Positive Rate (TPR), Accuracy (ACC), Sensitivity and Specificity parameters, and 5 output images respectively. Also, a qualitative comparison is made between MRF, HMRF and morphological image segmentation methods. Conclusion: The application of HMRF-Wavelet pattern recognition has been investigated in this work. VOR, RR, TNR, TPR, ACC, Sensitivity and Specificity are calculated, and most values have been reported as over 95% for five test images. Despite the complexity of the HMRF statistical modeling, acceptable performance measure values indicate the sufficiency of this algorithm.

Automatic sleep staging based on SVD, VMD, HHT and morphological features of single-lead ECG signal

Electroencephalogram (EEG) signals, which are among the primary polysomnography (PSG) signals used for sleep staging, are difficult to obtain and interpret. However, it is much easier to obtain and interpret electrocardiogram (ECG) signals. The use of ECG signals for automatic sleep staging systems could bring practicality to these systems. In this study, ECG signals were used to identify the wake (W), non-rapid eye movement (NREM) and rapid eye movement (REM) stages of the sleep data from two different databases with 17,758 epochs of 28 subjects (21 healthy subjects and 7 obstructive sleep apnea (OSA) patients) in total. Four different methods were used to extract features from these signals: Singular Value Decomposition (SVD), Variational Mode Decomposition (VMD), Hilbert Huang Transform (HHT), and Morphological method. As a result of applying the methods separately, four different data sets were obtained. The four different datasets were given to the Wrapper Subset Evaluation system with the best-first search algorithm. After the feature selection procedure, the datasets were separately classified by using the Random Forest classifier. The results were interpreted by using the essential statistical criteria. Among the methods, morphological method was the most successful and it was followed by SVD in terms of success rate for both two databases. For the first database, the mean classification accuracy rate, Kappa coefficient and mean F-measure value of the Morphological method were found as 87.11%, 0.7369, 0.869 for the healthy and 78.08%, 0.5715, 0.782 for the patient, respectively. For the second database, the same statistical measures were determined as 77.02%, 0.4308, 0.755 for the healthy and 76.79%, 0.5227, 0.759 for the patient, respectively. The performance results of the study, which is consistent with real life applications, were compared with the previous studies in this field listed in the literature.

Large Scale Dual Tree Complex Wavelet Transform based robust features in PCA and SVD subspace for digital image watermarking

A new watermarking scheme based on robust feature extraction using dual tree complex wavelet transform (DTCWT) in low dimensional subspace for grayscale imagesis proposed in this work. First the decomposition of host image is performed using DTCWT where large scale features having low frequency coefficients are extracted. The features are then further analyzed using the score matrix obtained from principal component analysis (PCA). Singular values from the score matrix are then obtained using the singular value decomposition (SVD) in a lower dimensional sub space. The watermark is subjected to similar processing after being scrambled using Arnold transform. The resultant robust lower dimensional DTCWT – PCA – SVD features are combined with the features extracted from the watermark. The robustness of the technique is evaluated by calculating quality assessment parameters. Further the robustness is evaluated according to the StirMark benchmarks by subjecting the signed image to various image processing attacks. High PSNR and correlation values as compared to state of art techniques indicate the imperceptibility and robustness of the proposed scheme.

Synergy of Schur, Hessenberg and QR Decompositions on Face Recognition

Human recognition through faces has elusive challenges over a period of time. In this paper, an efficient method using three matrix decompositions for face recognition is proposed. The proposed model uses Discrete Wavelet Transform (DWT) with Extended Directional Binary codes (EDBC) in one branch. Three matrix decompositions combination with Singular Value Decomposition (SVD) is used in the other branch. Preprocessing uses Single Scale Retinex (SSR), Multi Scale Retinex (MSR) and Single scale Self Quotient (SSQ) methods. The Approximate (LL) band of DWT is used to extract one hundred EDBC features. In addition, Schur, Hessenberg and QR matrix decompositions are applied individually on pre-processed images and added. Singular Value Decomposition (SVD) is applied on the decomposition sum to yield another one hundred features. The combination EDBC and SVD features are final features. City-block or Euclidean Distance (ED) measures are used to generate the results. Performance on YALE, GTAV and ORL face datasets is better compared to other existing methods.

Personnel recognition based on multistatic micro‐Doppler and singular value decomposition features

The use of micro-Doppler signatures experimentally collected by a multistatic radar system to recognise and classify different people walking is discussed. A suitable feature based on singular value decomposition of the spectrograms is proposed and tested with different types of classifiers. It is shown that high accuracy of between 97 and 99% can be achieved when multistatic data are used to perform the classification.

Classification of Unarmed/Armed Personnel Using the NetRAD Multistatic Radar for Micro-Doppler and Singular Value Decomposition Features

In this letter, we present the use of experimental human micro-Doppler signature data gathered by a multistatic radar system to discriminate between unarmed and potentially armed personnel walking along different trajectories. Different ways of extracting suitable features from the spectrograms of the micro-Doppler signatures are discussed, particularly empirical features such as Doppler bandwidth, periodicity, and others, and features extracted from singular value decomposition (SVD) vectors. High classification accuracy of armed versus unarmed personnel (between 90% and 97% depending on the walking trajectory of the people) can be achieved with a single SVD-based feature, in comparison with using four empirical features. The impact on classification performance of different aspect angles and the benefit of combining multistatic information is also evaluated in this letter.

Study on Hankel matrix-based SVD and its application in rolling element bearing fault diagnosis

Based on the traditional theory of singular value decomposition (SVD), singular values (SVs) and ratios of neighboring singular values (NSVRs) are introduced to the feature extraction of vibration signals. The proposed feature extraction method is called SV–NSVR. Combined with selected SV–NSVR features, continuous hidden Markov model (CHMM) is used to realize the automatic classification. Then the SV–NSVR and CHMM based method is applied in fault diagnosis and performance assessment of rolling element bearings. The simulation and experimental results show that this method has a higher accuracy for the bearing fault diagnosis compared with those using other SVD features, and it is effective for the performance assessment of rolling element bearings.

Roller Bearing Fault Diagnosis Based on ELMD and Fuzzy C-Means Clustering Algorithm

For the non-stationary characteristics of rotating machinery fault vibration signal, proposed a fault diagnosis method that based on ensemble local mean decomposition (ELMD) to extract fault feature, and fuzzy C-means clustering (FCM) to perform the fault identification. ELMD method can effectively solve the problem of aliasing modes in LMD. Firstly, decomposing the fault vibration signal by ELMD, PF components were obtained in which the initial feature vector matrix, The PF components compose a initial feature vector matrix, and do singular value decomposition, using the singular value decomposition feature vector as the fault characteristic vectors. Finally, using FCM clustering as a fault classifier. Achieved the identification of different fault types. Experimental results show that this method can effectively achieve the bearing fault diagnosis.

A SVD Feature based Watermarking Algorithm for Gray-level Image Watermark

In this paper, a novel gray-level image watermarking algorithm based on Singular Value Decomposition (SVD) feature and neural network is presented for copyright protection. Firstly, a new method which encodes the gray-level secret information into bit string is used. It uses the eigenvalues of SVD to extract the most valuable features of the watermark image and reduces lots of redundant information. Then based on human visual system's characteristic, the SVV (Singular Value and Variance) method is adopted to analyze the host image and choose the regions suitable for watermark embedding. Finally, the back propagation neural network is utilized for watermark recovery. Relying on the information contained in the extracted eigenvalues, the neural network is trained to recognize the original watermark. The application of the proposed scheme makes the gray-level watermark capable of transparent hiding. Results demonstrate that our method is robust to the intentional attacks and performs better than traditional methods.

Comments on “SVD-Based Modeling for Image Texture Classification Using Wavelet Transform”

This correspondence points out that in [1] there are some errors in two important formulas and the performance of the proposed singular value decomposition (SVD) feature is severely overestimated. It also presents a modified SVD feature which competes with the Gabor feature.

Singular value decomposition based feature extraction approaches for classifying faults of induction motors

This paper proposes singular value decomposition (SVD)-based feature extraction methods for fault classification of an induction motor: a short-time energy (STE) plus SVD technique in the time-domain analysis, and a discrete cosine transform (DCT) plus SVD technique in the frequency-domain analysis. To early identify induction motor faults, the extracted features are utilized as the inputs of multi-layer support vector machines (MLSVMs). Since SVMs perform well with the radial basis function (RBF) kernel for appropriately categorizing the faults of the induction motor, it is important to explore the impact of the σ values for the RBF kernel, which affects the classification accuracy. Likewise, this paper quantitatively evaluates the classification accuracy with different numbers of features, because the number of features affects the classification accuracy. According to the experimental results, although SVD-based features are effective for a noiseless environment, the STE plus SVD feature extraction approach is more effective with and without sensor noise in terms of the classification accuracy than the DCT plus SVD feature extraction approach. To demonstrate the improved classification of the proposed approach for identifying faults of the induction motor, the proposed SVD based feature extraction approach is compared with other state-of-the art methods and yields higher classification accuracies for both noiseless and noisy environments than conventional approaches.

Clinical detection and movement recognition of neuro signals

Neuro signal has many more advantages than myoelectricity in providing information for prosthesis control, and can be an ideal source for developing new prosthesis. In this work, by implanting intrafascicular electrode clinically in the amputee's upper extremity, collective signals from fascicules of three main nerves (radial nerve, ulnar nerve and medium nerve) were successfully detected with sufficient fidelity and without infection. Initial analysis of features under different actions was performed and movement recognition of detected samples was attempted. Singular value decomposition features (SVD) extracted from wavelet coefficients were used as inputs for neural network classifier to predict amputee's movement intentions. The whole training rate was up to 80.94% and the test rate was 56.87% without over-training. This result gives inspiring prospect that collective signals from fascicules of the three main nerves are feasible sources for controlling prosthesis. Ways for improving accuracy in developing prosthesis controlled by neuro signals are discussed in the end.