Recent Advances in Electrical & Electronic Engineering - Volume 9, Issue 1, 2016

Conventional thresholding shrinkage for denoising is designed with assumption that the coefficients in transformation domain are independent. However, in practice, natural images’ coefficients in transformation domain have significant dependencies. In this paper, we proposed a novel method for image denoising by exploring the dependencies among the coefficients. The method considered three corresponding coefficients, including the noisy coefficient, its parent coefficient and its neighbor coefficient based on the Sharp Frequency Localization Contourlet, and established a trivariate distribution model to estimate the latent coefficient. Furthermore, the shrinkage function in model is derived under the Bayesian framework. The experiment results showed that the performance of proposed method outperformed the current denoising method.

This paper focuses on studying an improved DR (Digital Radiography) technique in which the nonlinear transfer function and the adaptive unsharp masking method are coupled together. By introducing a product factor λ, this technique effectively wipes off the image noise and disturbance, while the interesting areas and borders in image are highlighted either. Besides, the whole enhancing process is automatic without loss in scalability. The experiment results show that this technique works well in most instances.

The traditional algorithm cannot simultaneously optimize the two indicators of high compression ratio and low information loss rate. To this end, a compression transmission method for multimedia image is raised based on improved neural network. Through the construction of image compression to transmit neural network mapping relation, and neural network was utilized to compress, transmit and reconstruct multimedia image, the learning process of neural network was improved also, by the introduction of game theory to make the corresponding parameters of compression and reconstruction self-adaptive change according to the results of the game, and study concurrently under the guidance of two objective optimization function, which improves the quality of image reconstruction after image compression and transmission. The experimental results show that, the improved algorithm can effectively improve the quality of reconstructed multimedia image after compression and transmission, the results is satisfactory.

The compression process of traditional image data compression method is complex, and has disadvantages like high computational complexity and low compression ratio. An image data compression transmission method based on singular value decomposition and wavelet transform is proposed and applied to multimedia network, the shortcomings of traditional method is analyzed, the singular value decomposition is applied to multimedia network image, and the energy ratio threshold is set, and relatively larger singular values are selected by self-adaption to complete the first stage compression of multimedia network image data. The threshold quantization method is utilized to process the high frequency coefficients under various resolution, the high frequency coefficients whose absolute value does not exceed the threshold is set to zero, so as to complete the image re-compression. The simulation results show that the proposed method can effectively enhance the image compression effect and reduce the complexity of the algorithm.

During the process of detecting crowd abnormality in sensitive multimedia image region, prone to severe occlusion. The detection process of traditional sensitive multimedia image region crowd abnormality based on cell movement characteristic, due to occlusion interference, the extracted motion feature is not accurate, thus, it is unable to obtain accurate crowd abnormality detection results, a sensitive multimedia image region crowd abnormality detection method based on behavior analysis is proposed, Hu invariant moment features as crowd abnormality feature is input into RBF neural network, through learning and classifying the RBF neural network, to access crowd abnormality of various types. Calculating and analyzing NMI of sensitive multimedia video image at each time moment, based on the analysis result of the crowd abnormality set a reasonable threshold, and complete the detection the crowd abnormality in sensitive multimedia image regions. The simulation results show that the proposed method has high detection accuracy.

Matching of fuzzy multimedia image has the disadvantage of larger error, a feature matching method for fuzzy multimedia image based on improved SIFT matching is proposed, on the basis of analyzing the shortcomings of the traditional method, multimedia images are preprocessed, and fuzzy space ideology is introduced in the algorithm to simulate all kinds of causes for fuzzy multimedia image, according to results to expand the database image set. Generating SIFT feature descriptors to determine the main direction of the feature points, the fuzzy image is rotated to the main direction and rotation invariance of the blurred image is maintained. Euclidean distance is used to measure the similarity of two vectors, through judging the ratio of the nearest neighbor distance and the second nearest neighbor distance exceeds a certain threshold, so as to determine the nearest adjacent points and the feature is the best match. The simulation results show that, the proposed method has high accuracy.

The edge part and noise of the fuzzy image have great uncertainty, the traditional fuzzy image region segmentation process based on fuzzy logic only considers the uncertainty problem, resulting in ignoring the complexity and correlation of fuzzy image information feature, unable to realize the fuzzy image segmentation efficiently (the paper is focus on how to achieve effective separation), a fuzzy image region segmentation method based on non-robust feature fitting is proposed, to analyze the disadvantages of traditional methods, and fit the non-robust features in the fuzzy image, based on the fitting results of non-robust feature, the optimal cluster number, clustering center and other parameters of the pixel are determined by fuzzy C mean clustering method, on the basis of pixels membership function to segment fuzzy image region. Through regional texture feature, spatial location feature, regional feature color and region shape character to describe the segmentation regions, so as to preserve the segmentation results, and provide basis for future fuzzy image region segmentation. The simulation results show that the proposed segmentation method has high accuracy, and a strong practical value.

This paper proposed a structure optimization principle according to the entropy calculating method of the HMM, and then introduced the TOM as the evaluation stander of the clustering result firstly. Then this paper proposed the WTOM by weighted the TOM, and divided the samples according to the value WTOM. Based on this work, this paper proposed an HMM based time-series images clustering algorithm which named HTICA, HTICA clustered the images hierarchically, theoretical analysis and experiments show that the proposed algorithm is very reliable, it can segment and identify the continuous time series samples unsupervised. And the application results in typical action extraction for badminton athlete is very well.

This paper presents an on-line inspection and warning system based on infrared image to detect the thermal fault of transformer bushings in the early stage. The network infrared thermal imager combined with the omnidirectional PTZ platform is the key device for the surveillance task. BRISK (Binary Robust Invariant Scalable Keypoints) algorithm is applied for the automatic object recognition and localization. In addition, the software based on B/S architecture is designed and realized for remote data transmission, so the remote users can acquire the real-time information through the private network. This system realizes the remote, real-time online monitoring and analyzing of transformer bushings, the comparison results validate the accuracy and efficiency of the recognition method, and the evaluations of the measured data verify the practicability of the system in the 110kV substation.

A design of CMOS class-F power amplifier (PA) at 2.4-GHz for wireless transmitters is presented. The class-F PA design is implemented by using 0.13-μm CMOS process. The proposed class-F PA employs cascade topology. The transistor’s on resistance is decreased by designing the transistors in parallel. Therefore, the efficiency is increased. The first stage is a common-source driver stage is biased in a class-AB to provide sufficient input voltage swing for the amplifier stage, while the amplifier stage is biased in cut-off region. Therefore, the transistor can operate as a switching-mode for high efficiency. The simulation results show that the power added efficiency (PAE) of 60% is obtained at 1.3 V power supply and the PA delivers 12 dBm output power. The chip area is 0.66 mm².

In this paper, an improved U-shaped microstrip meander-line slow wave structure (SWS) for low-voltage high-efficiency miniature G-band traveling wave tubes (TWT) is presented. To increase the TWT efficiency, the synchronization between electromagnetic (EM) wave and the electron beam should be maintained. Since the velocity of the electron beam passing along the tube is gradually reduced, the EM wave phase velocity should be also slowed down. Therefore, we gradually lengthen the width of meander-line to make longer the EM wave path for decreasing the EM wave phase velocity. Meanwhile, a simple analysis for calculating the dispersion properties of this kind of SWS is presented and the analytical results are compared with those obtained from simulation using CST-microwave studio. In addition, the scattering parameters of the SWS and the interaction between electron sheet beam and the electromagnetic fields are investigated with the help of CSTmicrowave studio and CST-Particle Studio, respectively. Our study shows that the G-band TWT is capable of delivering several tens of watts output power with 3-dB frequency bandwidth of 17GHz, ranging from 210GHz to 227GHz. Additionally, the peak output power is about 50 W with the corresponding gain of 40 dB at 218 GHz.

In this paper, a nonlinear single-cycle control method which adopts double closed-loop control has been studied on the single-stage AC/DC conversion device. The designed device could realize the functions of the stabilized voltage output, the improvement of power factor and the reduction of harmonic content, etc. First, one solution of the single-cycle control based on boost circuit topology has been proposed. Then, the main circuit and the control circuit are designed, respectively. Three control solutions are presented and the feasibility of them verification is compared via MATLAB. Finally, the single-cycle control solution is determined to be more compliant with the functional requirements of this design. The purpose of this study is to provide an overview about the single-stage AC/DC conversion device on the single-cycle control by addressing recent patents of such systems.