Earticle

To overcome the dependence on prior knowledge of traditional filtering algorithm, this paper proposes a novel non-parameter and adaptive multivariate empirical mode decomposition interval threshold (MEMD-IT) denoising approach for signal preprocessing of electromagnetic template attack(ETA). MEMD-IT can reduce the discontinuity induced by traditional MEMD-DT and aims to remove the Gaussian noise coupled into side-channel electromagnetic radiations. The proposed method and some other filters such as butterworth low-pass filter(BLPF)，wavelet threshold denoising and MEMD direct threshold (MEMD-DT) denoising are applied to analyze the electromagnetic radiation traces intercepted while cipher device was implementing RC4 encryption algorithm. Furthermore, twin SVM multi-class classifier based ETA was performed to evaluate the attack results. In the same attack scenario, the highest predictive success rate for 9 Hamming Weights of the key reached 92%,90%,84%,83% and 73% for MEMD-IT, WFTF, MEMD-DT,WULETF and BLPF preprocessing methods, respectively. The experiment results indicate that our proposed scheme has a significant performance compared with the traditional ETA.

The multiplier occupies more area in the design of FIR filter, which is the basic element of any filter. The low complexity and power architecture of pulse shaping FIR filter for multi standard digital up converter was designed. In the existing system, a 2-bit binary common sub expression (BCS) elimination algorithm and shift and add method are used to design an efficient multiplier in the pulse shaping filter. In this paper, two different types of adders are used in accumulation unit of the filter to reduce power consumption and delay compared to the existing system. Adders plays vital role in Digital signal processing applications and also used in the digital integrated circuits. With the advances in technology, several researchers have contributed towards designing adders with either high speed, less power consumption, low area or the combination of them. In this Paper, we describes the analysis of speed, power and delay of two different types of adders like carry select adder and carry save adder for accumulation unit of root raised cosine (RRC) filter. The number of additions are reduced by using these adders and succeeded in reducing the delay, power and area. The designed pulse shaping FIR interpolation filter is simulated and synthesized using Xilinx tool for Spartan 3E family devices and simulation results are presented.

Visual based location recognition of a mobile device is an important problem in many applications, such as visual navigation, auto-piloted driving and augmented reality. In this paper, a visual location recognition system based on the Coarse-to-Fine image retrieval and the epipolar geometry constraint is proposed. The basic idea of this system is to match a user captured image against some geo-tagged images in the database, and then estimate the user's location by the epipolar geometry constraint. The process of the Coarse-to-Fine image retrieval is necessary to select some database images in the same scene with the user captured image. The epipolar geometry constraint is utilized to determine the refined location using the geographical location information of the database images. The specific experiments for the visual location recognition are performed and the results show that this system can achieve the excellent performance of the location recognition.

Distribution Fields (DFs) for tracking achieved a better performance than traditional algorithms due to its special representation that allows smoothing the objective function without destroying information about pixel values. DFs descriptor can satisfy both the specificity and smooth landscape requirements of a good tracking algorithm. In this paper, we evaluate the Gray code in original DFs algorithm by replacing the pixel values using Gray code instead of original binary code. Experimental results show that the Gray code can improve the tracking efficiency in certain way.

In the present paper analysis of performance parameters i.e., insertion loss and return loss of microstrip Low Pass Filter with open stub using Artificial Neural Networks has been presented. The Artificial neural network is used in predicting the performance parameters of the low pass filter with open stub as a function of its stub length. Levenberg –Marquardt training algorithms of FFBP-ANN. (feed forward back propagation Artificial Neural Network), Layer Recurrent-ANN and CFBP-ANN (cascaded forward back propagation Artificial Neural Network) has been used to implement the neural network models. Simulated values for training and testing the neural network are obtained by analysing the LPF structure by the use of CST Microwave Studio Software. Comparison of mean square error obtained from different ANN networks concluded that CFBP-ANN gives satisfactory result as compare to FFBP-ANN and Layer Recurrent ANN. The testing of output of neural model is found good agreement with simulated output.

There are many disadvantages existing in the current experimenting platforms for embedded system course. Therefore, this paper has devised a hierarchical experimenting platform based on the CDIO education mode to improve both the teaching and learning experience. The experimenting platform devised in this paper is composed of a motherboard and several microprocessor plug-in boards. The motherboard can meet with various experimenting requirements by changing different plug-in boards. The demonstration source code is plug-in board independent, so the source code can be shared by different plug-in boards. The teaching practical shows that, this hierarchical experimenting platform can significantly improve the teaching and learning experience.

This paper presents a non-stationary object detection method by exploring time-varying spatial domain information in full motion video. Initially, the edge maps of difference image between two adjacent frames and current frame is generated via the well-known Canny edge detector. The distance of the edge pixels between the difference image and the current video frame are confined within a small value to determine the initial edge mask for the object in motion. The horizontal and vertical filling followed by morphological opening and closing operator are applied on the initial edge mask to create initial temporal segmentation mask of the moving object. The morphological dilation and corrosion operator are utilized to obtain binary marker image of the foreground and background which are used to modify the multi-scale morphological gradient image of current frame. Finally, the watershed algorithm is performed on the modified gradients to find the non-stationary objects accurately in the spatial domain of motion frames. Processed video results show detection accuracy of 98% and 99% for four different video experimentation test-beds involving fast and slow human motion. In this operation, the proposed technique eliminates the problem of over-segmentation of the watershed algorithm and extracts visually distinct, contextually meaningful non-stationary objects as they randomly appear (or disappear) in video sequences.

On-orbit servicing robots were developed to complete the missions in the uncertain and hazards outer space. The operator controls the remote robot via teleoperation systems, which allows the operator to manipulate the robot at a distance. But, as the existing of communication time-delay affects the whole system stability, transparency and other performance index. Consequently, a proper teleoperation system could help the operator to perform missions more accurate. This paper presents the general scheme of the teleoperation system, compares the four typical control modes and estimates methods of time-delay. At last provides an outlook to future research directions.

The segmentation results of the traditional FCM based image segmentation algorithms are only determined by the distribution of pixel intensity in the feature space, and they does not take the spatial distribution of pixels into consideration, which make the segmentation results discrete in the spatial distribution. To solve this problem, a global spatial similarity metric and a global intensity similarity metric are proposed, and introduced to a new distance metric which is used to calculate the difference between pixels and cluster centers. In addition, a maximal similarity based class merging mechanism is employed to achieve more accurate image segmentation. The experiments demonstrate that, comparison with the FCM and KFCM based image segmentation algorithms, the proposed method produces more accurate and applicable segmentation results.

Speaker Recognition and Verification is becoming one of the widely used forms of biometric authentication in today’s scenario where remembering strings of textual passwords and numbers are becoming a hassle. Authentication of users using voice offers many advantages and easy to use techniques. In this paper a comparison is drawn among the most commonly used feature extraction techniques in Speaker Recognition and Verification. Extracting useful and unique features from the user’s voice forms the backbone of an efficient Speaker Recognition System. Here, the most commonly used methods for Feature Extraction viz. MFCC (Mel Frequency Cepstral Coefficient), LPC (Linear Predictive Coefficient), PLP (Perceptual Linear Prediction) are discussed, compared and an attempt is made to deduce which one performs best.

Grading the fruit by using machine vision system, it is hoped by the computer to recognize and understand the image automatically, in order to achieve this objective, the key step is to capture the suitable fruit images so that fruit image information can be effectively decomposed. Therefore, the final result of decomposition is to get some of the characteristics of each image with its own motifs, such as borders, shape and so on. By using these primitives, you can match a certain pattern, so as to determine the quality of the fruit. In this paper, it takes the overview of the finite element segmentation as a starting point, combined with the interpretation of the numerical algorithm and FCM algorithm functional convergence of the sequence, relying on Mumford-Shah function model to investigate the generation of fruit image finite element model.

To meet the demand of rapid, accurate and online inspection of precision parts’ defects, the CCD camera is used to collect information of the inner hole. Since the obtained source image is a concentric annular image of 360 degrees with large image distortion, meanwhile the collected source images are not suitable to be processed directly with the image processing technology. So further processing is needed to be done for the rectangle unwrapping images. The method of Hough transform is adopted in this paper to find the center of a circle and the corresponding position used the center as a reference. On the basis of positioning the circle center, a kind of circular image unwrapping algorithm is designed by the transformational relation between Cartesian and polar coordinate. During unwrapping of the image, the bilinear interpolation algorithm is adopted to improve the image quality when the image unwrapping algorithm is used. The algorithm is verified by program and the result indicates that the algorithm to unwrap the image is efficient in locating the center by Hough transform combined with the coordinate’s transformation and the bilinear interpolation. Furthermore, the algorithm can unfold the image accurately and quickly and ensure the quality of the image, which prepares the foundation for subsequent processing.

This article describes research on the topic of the influence of skin diseases on fingerprint recognition, particularly the actual detection of fingerprint areas negatively influenced by skin diseases. It introduces a number of skin diseases, describes their influence on fingerprint images according to a diseased fingerprint database, explains methods used for the detection and finally presents results of this research.

The classification accuracy is an important standard to measure the quality of the classifier. Usually, the classification accuracy is assessed later, not during the classification process. Problems such as classification accuracy drops cannot be timely and effectively found. It is necessary that marking test samples when estimating classification accuracy. It is a problem that we care about that how much is the classification accuracy when a group of new samples obtained. The problem must be concerned when using and improving the classifier in the case of growing data. To solve this problem, this paper put forward different estimates from different perspectives which based on the difference between samples. One estimate is based on the difference in samples distribution, which is from the Bayesian criterion. Another estimate is based on the difference in each sample instance, which is from the K nearest neighbor classification. Classification accuracy is also estimated by using the artificial neural networks, which combine the characteristics of the above two methods. And results show the proposed methods have good effects.

In this paper Application of an Empirical Wavelet Transform based technique is proposed to estimate time-varying PQ indices for accurate assessment of Power Quality Disturbances. The EWT approach mainly aims to extract the actual fundamental frequency component and disturbance components from any distorted signal. The empirical wavelet transform consists of two major steps: detect the Fourier supports, and build the corresponding wavelet accordingly to those supports; filter the input signal with the obtained filter bank to get the fundamental component and disturbance components. Since the extracted components contain only one frequency component, Hilbert transform is utilized to estimate the instantaneous frequency and amplitude information, from this information we can estimate time-varying PQ indices. The proposed method is employed to assess successfully all sorts of Power Quality Disturbances such as voltage sag, swell, interruption, transients, harmonics, spikes, notches etc. From the results we can say that the proposed method detects disturbance start time, end time, duration of existence and its content more accurately.

The maize seeds recognition is an important aspect of the seed quality testing and breeding. With the development of computer technology, it is an inevitable tendency to recognize seeds varieties and inspect quality using machine vision instead of human vision. For the characteristics of maize seeds and their images, seeds varieties recognition technology and algorithm was studied and explored in depth and their shape and color feature parameters were extracted. Based on 4 species including Nongda 108, Zhengdan 958, Ludan 981 and Jingdan 28, the classifier was designed using BP neural network to do varieties recognition testing. Research showed the average recognition accuracy was more than 94.5%, and that recognition and detection of maize seeds varieties based on the image technology is feasible.

The image segmentation technology is of great significance to the target identification. The watershed segmentation algorithm has wide application in image segmentation. The traditional watershed segmentation often causes the problems of over segmentation and noise sensitivity. Therefore, a medical image segmentation algorithm is proposed based on K-means clustering algorithm and improved watershed algorithm. First, K - means clustering algorithm is used for initial segmentation, and then the concept of similarity is put forward to improve the original watershed algorithm. Finally, the adjacent tiles of the initial segmentation is merged. The magnetic resonance image is regarded as the segmentation object. The experimental result shows that the proposed algorithm effectively solves the problem of the over-segmentation of traditional watershed algorithm, and achieves a satisfactory effect for the image segmentation.

This document signifies the overview of the Unconstrained Face Identification Quality within an unconstrained face-recognition atmosphere. It offers the functions like imbalance, lighting, and phrase variance strategy. It may be employed for both confirmation and id. At the moment, there are lots of means of front watch face-recognition. Couples of years ago, for computer-vision, numerous face-recognition methods were prepared. But, real-globe face-detection quiet deposit a challenging work. The interest in unconstrained useful face-recognition is growing using the detonation of online media for example neighborhood systems, and video-surveillance video wherever encounter evaluation is of great significance. Within this evaluation, we show up to manage reputation within the situation of graph presumption. We are able to identify a mystical encounter utilizing a Diverse Method. This research lights-out the options recommended for unconstrained face recognition quality region and recommending the clear answer to become used by RFG (Reference Face Graph) based face-recognition. Our purpose is RFG centered unconstrained face-recognition to improve the demonstration quality. The Formula and Simulator of recommended sensing methods is likely to be completed by utilizing Mat-laboratory.

As the reform and opening up going into depth over the past three decades and more, the market economic system has been gradually established. The banking industry grows steadily in the process of the reform. It supports economic development, reduces and defends many financial risks in the processof the reform. However, there are many kinds of risks inside of banks, one of which is that the non-performing loans (NPLs) ratio is too high. Therefore, people should focus on how to accurately classify the banking loans into performing and non-performing ones and how to control and prevent the resulting crisis. This paper deeply analyses China’s NPLs problem for the current period, recognizes and classifies loans types by adopting decision trees, Naïve Bayes and support vector machine (SVM) methods. The experiment result found that the decision trees method can well identify the performing loans and non-performing loans; its accuracy is as high as 94%.

In order to solve the problem of image reconstruction in electrical capacitance tomography (ECT) technology, the feasibility of applying convolutional neural network (CNN) to ECT image reconstruction is studied. The convolution layer and the training of the structure of the sub sampling method is improved based on deep research for convolution neural network for the more time-consuming process of deep structure and training issues, and a fast convergence convolution neural network (FCCNN) image reconstruction method is proposed. Matlab was used to build a ECT simulation system. For each algorithm, the simulation results were compared and analyzed. The experimental results show that our algorithm improved the image reconstruction efficiency and quality of the common flow pattern.

Now a day’s optical communication is a major technology that meets end user demands. In fiber optical communication Wavelength Division Multiplexing (WDM) technique is introduced and it supports adequate bandwidth for the end users. The route assignment of WDM is an important in case of optical routing networks. Here in this paper we provide the solution for the problem of fault detection for routing and wavelength assignment that meets the end user requirements for increasing the efficiency of wavelength routed All-optical network traffic.

A novel local region-based active contour model is proposed to segment medical images with intensity inhomogeneities and various noises. The contribution of the proposed work is twofold. First, the anisotropy of evolution contours is exploited to characterize the local classification information around each pixel. Integrating it with local gray intensity information, the new model stabilizes the active contours in all evolving processes. Second, under the constraint of maximum absolute error of parameter estimation, the optimal spatial scales are automatically selected for the local segmentation models. It is demonstrated from the experiments that our algorithm achieves faster and more robust results than several same-type methods.

The 3D data acquisition system is a very effective means to protect handicrafts. This paper analyzed the advantage of 3D laser scanning technology and designed a 3D data acquisition system on handicrafts base on it. The hardware of this system includes 6 parts: laser range finder, computer, servo system, interface board, distance data instrument, angle data instrument. The software of this system have 4 functions: parameter setting, automatic scanning control, scanning data processing, fault diagnosis. This paper discussed the operation of the software platform and development tools, modular design of software, design of software system framework and information fusion of 3D data acquisition. This paper completed the fusion and matching of point cloud data.

Acoustic Emission (AE) source localization is a useful implement to diagnose the incipient faults in rotating machine. This paper proposes an improved Near-field Multiple Signal Classification method using four-order cumulants (NFC-MUSIC) to accurately locate the faults position. In order to overcome dispersion effect and revised velocity, the feature sub narrow band is extracted by Wavelet Packet reconstitution (WPR) and Modal Plate Wave Theory (MPWT). For multi-source decorrelation and increasing localized resolution, the four-order cumulants of observed signal can be selected for localization. The experiment results indicate that the improved method can accurately locate multi-rubbing faults. It is an efficient way to assist incipient rubbing faults diagnosis.

In recent years the Wavelet Transform (WT) had an important role in various applications of signal and image processing. In Image Processing, WT is more useful in many domains like image denoising, feature segmentation, compression, restoration, image fusion, etc. In WT based image fusion, initially the source images are decomposed into approximation and detail coefficients and followed by combining the coefficients using the suitable fusion rules. The resultant fused image is reconstructed by applying inverse WT on the combined coefficients. This paper proposes a new adaptive fusion rule for combining the approximation coefficients of CT and PET images. The Excellency of the proposed fusion rule is stamped by measuring the image information metrics, EOG, SD and ENT on the decomposed approximation coefficients. On the other hand, the detail coefficients are combined using several existing fusion rules. The resultant fused images are quantitatively analyzed using the non-reference image quality, image fusion and error metrics. The analysis declares that the newly proposed fusion rule is more suitable for extracting the complementary information from CT and PET images and also produces the fused image which is rich in content with good contrast and sharpness.

Aiming at the problems that the luminance information is not enough and the edge information can not be preserved easily in the process of the image fusion, an effective image fusion method based on nonsubsampled contourlet transform (NSCT) and adaptive guided image filtering is proposed in this paper. Firstly the source images are decomposed by NSCT with multi-scale, multi-direction and shift-invariant properties. The fusion rule of the low-frequency subband coefficients employs the local correlation energy to improve the energy and information of the subband coefficients. To acquire good effects with edge-preserving and noise reduction, the adaptive guided image filtering is introduced to the high-frequency subband coefficients as the fusion rule for the first time, and it can make use of halo-free edge slope enhancement in the fusion process. Finally, the fused image is reconstructed by the inverse NSCT. Experimental results demonstrate that the comprehensive performances of the proposed method are improved in the fusion information, edge and luminance.

In this paper, model-based design is used to complete the design of an adaptive filter by Least Mean Square (LMS) algorithm, which implements the recovery process of audio signal disturbed by noise. We can quickly build a system simulation model by model-based design approach, and accomplish efficiently the system test, simulation and implementation. Theoretical analysis and experimental results show that the method of model-based design is not only valuable to the design and implementation of DSP system, but also can significantly improve the design efficiency of the DSP system.

Aiming at the traditional GEP algorithm adopted fixed rate of mutation and crossover rate in the process of evolution, and ignored the dynamic change of individual fitness, which leaded to the presence of premature convergence and local optimization problem. By using the cloud adaptive strategy and cloud cross strategy of cloud model, a genetic algorithm based on cloud model (Cloud Model Gene Expression Programming, CMGEP) was proposed. The algorithm adjusted the mutation rate and crossover rate in evolution through the cloud adaptation strategy according to the change of dynamic, and timely calculated population similarity to achieve cloud cross to increase the diversity of population and jump out of the premature convergence. It was applied to the field of railway engineering and its results were compared with those obtained by traditional GEP Algorithm and CMGEP Algorithm. Experiments show that the algorithm can improve the adaptability and the prediction accuracy, it has better convergence.

We are intended to propose a video compression based on Modified Pixel-Based Motion Estimation (MPME) technique to solve the current problems in the literary works. The proposed system will be comprised of four phases, Displacement Calculation, Motion Estimation, Encoding and Decoding. In displacement calculation the I-frame and the P-frame has to been constructed from the video frames. In second phase, the motion vector has been finding out by using the Modified Pixel Based Motion estimation. In here we are creating four movement vector predictors (West Predictor, Neighbor Predictor, Median Predictor and Centre Predictor) with the help of diamond based search pattern. The proposed motion estimation technique is conducted on a pixel-by-pixel basis and the predicted P-frame has been developed using motion estimation technique. The residual is calculated with the original P-frame and the predicted P frame, and the residual is passed into next phase. After the motion vector estimation each frame is encoded by using JPEG-LS encoder. Next the encoded values of compressed video frames are decoded by using the decoding technique and for that JPEG-LS decoder is used. The proposed video compression using MPME will be implemented in the platform MATLAB and the performance of the video codec procedure is assessed by contrasting the consequence of proposed system with the conventional motion estimation algorithms.

This paper studies the optimal linear estimation problem for systems with multiple packet dropouts and multiplicative noises. When the current measurement is lost, the popular zero-input compensation mechanism is used for compensation. Based on the zero-input compensator, the optimal linear estimators including filter, predictor and smoother in the linear minimum variance sense are given by innovation analysis approach. The proposed estimators can reduce the computational cost compared to the existing augmented estimators based on the hold-input mechanism. The performances of the two kinds of estimators are compared in terms of two simulation examples. The conclusion is that neither of the two compensation mechanisms can be claimed to be superior to the other.