Earticle

In recent years, tensor completion problem has received a significant amount of attention in computer vision, data mining and neuroscience. It is the higher order generalization of matrix completion. And these can be solved by the convex relaxation which minimizes the tensor nuclear norm instead of the n-rank of the tensor. In this paper, we introduce the weighted nuclear norm for tensor and develop majorization-minimization weighted soft thresholding algorithm to solve it. Focusing on the tensors generated randomly and image inpainting problems, our proposed algorithm experimentally shows a significant improvement with respect to the accuracy in comparison with the existing algorithm HaLRTC.

Spread Spectrum communication system is one of the most important types of the modern telecommunication systems. It involves diverse activities and processes where time restrictions are very important which formulate these systems very complex. To solve the complexity of these systems and to process the involved activities in time, the Real Time System (RTS) concept, with a vital role, came into being. RTS is an information processing system which has to respond to some input stimuli with in finite & specified period of time and to schedule the shared resources among different tasks. This paper examines that how real time system applications are involved in spread spectrum communication.

Biometric authentication has been widely studied for many years and attracted much attention due to its large ability security application. Palm vein is more immovable and more difficult to fake than other biometrics such as fingerprint, palm print and face. Since palm veins exist inside of the body, it is exceedingly hard to be forged. Palm vein authentication uses the unique patterns of the palm vein to identify individuals at a high level of accuracy. In the proposed work, the palm vein image enhancement algorithm proposed based on Gaussian matched filtering and then two types of feature extraction are extracted. The global features based on wavelet coefficients and locale feature based on local binary pattern (LBP). In the propose work, a linear dimensionality reduction algorithm, called Isometric Projection is used. Finally, the Manhattan Distance (MHD) matching method is proposed to verify the test palm vein images. The experimental result shows the EER to the proposed method is 0.17488%.

This paper presents a hybrid content-based video coding scheme that encodes arbitrary shaped objects instead of blocks of images. The scheme achieves efficient compression for low bit-rate applications by separating moving objects from stationary background and transmitting the shape, motion and residuals for each segmented object. Furthermore, a new content-based object segmentation algorithm is proposed in the scheme, which does not assume any prior modeling of the objects being segmented. The algorithm is based on a threshold function that calculates block histograms and takes image noise into account. The experimental results show that the scheme proposed outperforms the classical object-based coding methods in terms of PSNR or the average number of bits required for coding a single frame.

A novel color image compression scheme based on block truncation coding is proposed in this paper. In this scheme, the quadtree segmentation technique is employed to partition the color image into variable-sized image blocks according to its block activities. Then, the block truncation coding scheme with the bit map omission technique is employed to encode the image blocks. According to the experimental results, the proposed scheme significantly cuts down the bit rates while keeping good reconstructed image qualities of the compressed color images.

This paper utilizes a 0 Ohm substitution current probe to measure the emission on the power network of an integrated circuit (IC). The measurement result shows that current probe works up to 1GHz, and the short time FFT (STFFT) method is used to identify the emission source. All the wideband emissions are from DC/DC converter, and no narrowband emissions are found. The measurement method gives great convenience to measure the current emission on printed circuit board (PCB) trace.

The color filter array (CFA) captures only one-third of the necessary color intensities and the full color image is generated from the captured data by interpolation. In recent years, the algorithm of Bayer patterned image compression based on “structure separation” has achieved better image quality. On the basis of previous work, the algorithm based on the all phase biorthogonal transform (APBT) and interpolation is proposed in this paper. Instead of the conventional DCT-JPEG, the APBT-JPEG significantly reduces complex multiplications and makes the quantization table easier. Several kinds of interpolation methods to the decompressed image data are also discussed in this paper, including nearest neighbor interpolation, bilinear interpolation, cubic convolution interpolation and a novel interpolation method based on APIDCT. Experimental results show that the proposed algorithm outperforms the one based on “structure separation”; and the APIDCT interpolation performs close to the conventional interpolation methods and behaves better than them at high bit rates.

To get better and flexible performance, region-based image coding has been studied widely in the last few years. In region-based image coding, the whole image can be divided into two parts: region-of-interest (ROI) and background area. ROI is coded in low compression ratio, while background area is coded in high compression ratio. In this paper, shape-adaptive DCT (SA-DCT) for coding arbitrarily shaped image segments is introduced. And two region-based image coding methods using discrete cosine transform (DCT) and SA-DCT are presented. The algorithm based on SA-DCT is comparable with the one based on DCT in terms of reconstructed images’ effects. Experimental results show that the algorithm based on SA-DCT performs better than the one based on DCT both in subjective effect and objective effect. Therefore, SA-DCT is a more suitable transform for region-based image coding than DCT.

Speaker Recognition is a challenging task and is widely used in many speech aided applications. This study proposes a new Neural Network (NN) model for identifying the speaker, based on the acoustic features of a given speech sample extracted by applying wavelet transform on raw signals. Wrapper based feature selection applies dimensionality reduction by kernel PCA and ranking by Info gain. Only top ranked features are selected and used for neural network classifier. The proposed neural network classifier is trained to assign a speaker name as label to the test voice data. Multi-Layer Perceptron (MLP) is implemented for classification, and the performance is compared with the proposed NN model.

The accurate and rapid identification of landslide region is the basis for emergency disaster processing and analysis. The identification methods based on change information among multi-temporal remote optical sensing images are simple and intuitive. But its performance is remarkably limited by issues, such as unable to distinguish the changes according to the causes, improper processing strategy for impact of the adjacent pixels and the low efficiency. Thus, an approach based on digital elevation model and change information is presented in this paper. At first, the potential landslide region is calculated through slope information. And then the change information constrained by aspect information is used to identify and extract the final landslide regions. The experiment results show that, this approach can effectively distinguish the change information caused by city construction or landscaping planning. The extracted landslide regions are greatly consistent with interpretations. Meanwhile, it is effectively satisfy the demands for emergency disaster processing.

The detection of breast cancer in its early stage is a developing research field that can rescue the women affected by cancerous tissues. A modest and skilful approach in microwave imaging (MWI) for data inversion is proposed. MWI in medicinal applications has been famous in last two decades and presently it is using for the detection of breast cancer. Medical imaging by microwave technology has been potential to investigate the electrical distribution, dielectric properties, and as well conductivity for malignant tumor inside the breast. In this study subscription, the writers have reviewed the research importance for inverse scattering method in the microwave tomographic imaging as well as discussed Born type approximation for malignant tumor detection. Distorted Born iterative method (DBIM) will also be discussed. This paper presents the approaches of image reconstruction for breast cancer detection.

The saturation characteristic of the core varies with operation of magnetically saturation controlled reactor’s (MSCR). Consequently its harmonic characteristic is needed to take in consideration. The arithmetic representation of MSCR is derived in accordance with the structural characteristic and operational rules. The characteristic analysis investigates the function of voltage as well as current over all magnetic field parameters of harmonics. After that, it concludes that the working current of MSCR’s is an odd harmonic function, including primary and odd harmonic components; controlling current and voltage are even harmonic functions, restraining DC and even harmonic components; Again magnetic point of view, the magnetic components are also surrounding the DC and its own fundamental part too. Finally, a simulation model for MSCR is fabricated with MATLAB/Simulink. This paper demonstrates that the analysis is very precise. So, this paper can be a reference to provide further instruction for analysis and proposition of new harmonic suppression methods of MSCR.

Color image segmentation algorithms are proposed based on granular computing clustering (GrCC). Firstly, the atomic hyperspherical granule is represented as the vector including the RGB value of pixel of color image and radii 0. Secondly, the union operator of two hyperspherical granules is designed to obtain the larger hyperspherical granule compared with these two hyperspherical granules. Thirdly, the granular computing clustering is developed by the union operator and the user-defined granularity threshold . Global Consistency Error (GCE), Variation of Information (VI), Rand Index (RI), and Loss Entropy (ΔEn) are used to evaluate the segmentations. Segmentations of the color images selected from internet and BSD300 show that segmentations by GrCC speed up the segmentation process and achieve the better segmentation performance compared with Kmeans and FCM segmentations.

A method based on local contour distribution features is proposed for writer identification in this paper. In preprocessing, contours are abstracted form images by an improved Bernson algorithm. Then the Local Contour Distribution Feature (LCDF) is extracted from the fragments which are parts of the contour in sliding windows. In order to reduce the impact of stroke weight, the fragments which do not directly connect the center point are ignored in the feature abstraction procedure. The edge point distributions of the fragments are counted and normalized into LCDFs. At last, the weighted Manhattan distance is used as similarity measurement. The experiments on our database and ICDAR 2011 writer identification database show that the performance of the proposed method reach or exceed those of existing state-of-art methods.

Multi-target tracking based on on-line boosting is a significant technique in computer vision. However, it is very difficult to select the optimal classifier in on-line learning process since tracking often relies on an assumption that the appearance model of target is fixed. This would directly lead to a decline in the performance of on-line boosting. In this paper, we presents a novel on-line multi-target tracking framework based on Weighted Incremental Histogram Model (called WIHM), which can be applied in some static or dynamic scenarios. First, we propose a novel method—WIHM, which is employed to obtain the optimal size of a tracked object. Second, a new update scheme is used to reach local optimum tracking appearance model with high possibility and accuracy. Third, based on the above works, a multi-target tracking framework is proposed to track multi-targets simultaneously. With the appearance model of targets changed continually, our represented approach can track these targets more powerful, especially in dealing with camera motions. Experimental results show the effectiveness and robustness of our method.

Speckle Noise is a natural characteristic of medical ultrasound images. It is a term used for the granular form that appears in B-Scan and can be considered as a kind of multiplicative noise. Speckle Noise reduces the ability of an observer to distinguish fine details in diagnostic testing. It also limits the effective implementation of image processing such as edge detection, segmentation and volume rendering in 3 D. Therefore; treatment methods of speckle noise were sought to improve the image quality and to increase the capacity of diagnostic medical ultrasound images. Such as median filters, Wiener and linear filters (Persona & Malik, SRAD ... ..).The method used in this work is 2-D translation invariant forward wavelet transform, it is used in image processing, including noise reduction applications in medical imaging.

A microcalcification detection method based on wavelet singularity was presented because of microcalcification singularity characteristic. Firstly, the source image is decomposed in multi-scales wavelet coefficients. Secondly, coefficients in low-pass band are removed and coefficients in high-pass band are enhanced contrast by nonlinear method. Lastly, fisher discriminant was adopted in segment microcalcifications. Experiment results showed that wavelet basis with shorter support and lower regularity is more sensitive to noise, while wavelet basis with longer support, higher regularity and higher order vanishing moment could segment indistinct microcalcifications, but sometime could not segment small microcalcifications. The results also showed the detect effect DAUB4 wavelet is best and its detection ratio is about 96%.

Region-based level set segmentation is a paradigm for the automatic segmentation of brain tumor image. Unfortunately, region-based segmentation, which is relied on the intensity difference of different regions, has been of limited used in presence of complex background. In fact, the evoluting curve may leak out the boundary of tumor to reach a steady state by the global region force. In this work, we propose a new hybrid approach for brain tumor segmentation, which is relied on the approach of global intensity difference, local edge properties, curve evolution, and level set method. The regional information drives the contour to converge to the global minimum. By combining the edge information into the region-based framework, the images with intensity inhomogeneity and complex background can be efficiently segmented. To improve the accuracy of brain tumor segmentation, a skull-stripped method for brain images is proposed by utilizing the new morphological process. In addition, a penalizing energy is used for avoiding the time-consuming re-initialization step of the level set method. Finally, experiments are preformatted on some synthetic and real images. By visually assessments, results on patients demonstrate the new method can segment tumors with few iteration times. Moreover, comparisons with the most similar methods also show that the proposed method is effective for the segmentation of tumor in MR image.

In this paper, face recognition using the most representative SIFT images is presented. It is based on obtaining the SIFT (SCALE INVARIANT FEATURE TRANSFORM) features in different regions of each training image. Those regions were obtained using the K-means clustering algorithm applied on the key-points obtained from the SIFT algorithm. Based on these features, an algorithm which will get the most representative images of each face is presented. In the test phase, an unknown face image is recognized according to those representative images. In order to show its effectiveness this algorithm is compared to other SIFT algorithms and to the LDP algorithm for different databases.

Multiple fractal dimensions can be used to depict the geometry characteristic of the radiation source signals from different dimensions, thus it can be used to extract the features of different radiation source signals. In this paper, it proposes an improved feature extraction algorithm of radiation source signals based on the multiple fractal theory, it improves the solution method of traditional multiple fractal dimension which accumulated the q dimensions characteristics. Meantime, it increases the regularity and gathered degrees of the characteristics of radiation source signals, and under the condition that the basic computational complexity of algorithm is not changed. Simulation results show that, for the classification of different radiation source signals, the improved algorithm has better property and classification rate.

The interest and focus for quite some time has been on Feature Selection and lot of work has been made in this field. With databases getting larger in volume so machine learning techniques are required which results in demand for feature selection. Feature selection is commonly used method for performing data mining in the field of data preprocessing that is scaled on large amount of data sets. In this paper, several kinds of feature selection methods are used which may result in different subsets of features with evaluation criterion.

"Soft-field" nature and ill-posed problems to influence on the accuracy of image reconstruction in electrical capacitance tomography technology (ECT), in the analysis of ECT basic principle and imaging algorithm, a novel image fusion method based on edge feature detection using the feature of different frequency domains of wavelet decomposition is presented in this paper. The LBP algorithm and Landweber algorithm are used as the base of image reconstruction in the paper, the fusion rules based on the combination of Local gradient and local signal strength and the weighted averaging operator are used to fuse the high-frequency coefficients and low-frequency coefficients of the fused image, in this way , higher accuracy imaging results can be got. The simulation results show that the image fusion accuracy is improved, and reconstruction image effect is enhanced.

This paper presents musical instrument recognition for isolated music sound signals using hybridization of fractional fourier transform (FRFT) based features with timbrel (acoustic) features using feed forward neural network. The FRFT based features which is named as fractional MFCC are computed by replacing conventional discrete fourier transform in mel frequency cepstral coefficient (MFCC) with discrete FRFT. Hybrid features are obtained by effectively combining Fractional MFCC with timbrel features such as temporal, spectral and cepstral features. Feed forward neural network with back propagation algorithm has been used to test the performance of system and results were compared in terms of recognition accuracy and number of features. Proposed feature out performs over individual and other traditional features proposed in the literature. The experimentation is performed on isolated musical sounds of 19 musical instruments covering four different instrument families. The system is tested on benchmarked McGill University musical sound database.

The High resolution (HR) images can be obtained from a set of noisy and blurred low resolution (LR) observations by applying the Super Resolution (SR) technique. In this paper a new SR algorithm that uses Singular Value Decomposition (SVD) based Fusion and Blind deconvolution techniques is proposed. The algorithm significantly improves the resolution and eliminates the noise and blur associated with low resolution images, when compared with the other existing methods.

In the communication mechanism of co-time co-frequency full-duplex (CCFD), digital self-interference cancellation (SIC) is used for suppressing residual self-interference (SI) after antenna and radio frequency (RF) SICs, as well as low-power multipath SI. In this study, the CCFD LTE verification experiment is presented, which adopts digital SIC with the SI signal reconstructed in frequency domain. For the multipath Rayleigh SI channel, the expression of digital SIC capability is derived and the relationship among the channel estimation error, the received SI power, and digital SIC capability is analyzed. Experimental results show that the digital SIC ability is 31.2 dB for a 20 MHz 16QAM modulated LTE SI signal with frequency centered at 2.6 GHz and the interference-to-noise ratio of 40 dB.

The paper mainly studies static 2D face images through reconstructing 3D model by a specific algorithm. First, the paper need collect geometric features, and obtain the three-dimensional space of false geodesic distance. Those are to emotional changes. Second, remove the relative feature extraction. Finally, compares the test sample and the training samples about the Mahalanobis distance. The experimental results show the recognition rate reaching to ninety percent. Results illustrate the validity of the algorithm.