Applied Mechanics and Materials Vols. 548-549

Abstract: This paper discusses the various texture-based and vector-based approaches to classify coins.These two types of approaches are common for software-based coin sorting systems. Many researchers have applied algorithms known in artificial intelligence research for feature extraction, selection and classification of coin images. However, without a common benchmark, it is difficult to assess the accuracy, robustness and efficiency of the coin sorting systems across the different approaches. It is proposed that the use of a standard benchmark image databank such as the CIS Benchmark will allow a more objective and accurate comparison of the performance of these coin classification approaches.

Abstract: We present a novel approach called Global Color Composition (GCC) to improve the visual effect of large image show on curved screen multi-projector display system. Instead of traditional approach only blend overlapped area, our algorithm utilize the global gradient-domain information to get better effect, and as an optimizing of the algorithm to reduced time and memory, we utilize the patches border pixels to construct the passion equation. The solver of this passion equation is result image we desired, which can be solving on parallel GPU threads. Experiment results are given to illustrate the effect of our approach.

Abstract: Graph indexing is one of the budding dominant current research domains. Varieties of data from different domains pop into the daily updating world in huge volume. There are plenty ways to analyze and to arrange them into an easily accessible index. Here in this paper, we concentrate on the graphical approach. The complete classical algorithms of graph indexing are discussed initially followed by the latest proposed solving methods. Alas, at the end of the paper a complete PLANO frame work is proposed such that the given data can be easily indexed and the updated features can be propelled onto the index.

Abstract: Compressed sensing (CS) theory breaks through the limitations of the traditional Nyquist sampling theorem, and accomplishes the compressed sampling and reconstruction of signals based on sparsity or compressibility. In this paper, CS theory is used to do the parameter estimation of wideband Linear Frequency Modulated (LFM) signal in order to decrease the sampling pressure. A novel method that reconstructs the edge information of the LFM spectrum based on wavelet transform and CS theory is proposed. On the basis that the wideband LFM signal has approximate rectangular spectrum, the wavelet-based edge detection is introduced to provide sparse representation for the signal spectrum. The edges of the spectrum can be reconstructed by the CS reconstruction algorithms. Consequently, the initial frequency and final frequency of wideband LFM signal can be estimated with high estimation precision. The effectiveness of the proposed method is confirmed with numerical simulation.

Abstract: We firstly propose a hybrid-Radio-and-Visible-Light-Communication (HR-VLC) system in Inter-Vehicle Communication (IVC) employing Spread-Spectrum (SS) for better ranging and larger capacity. IVC, a key in Intelligent Transport System (ITS), is intended to improve the better safety and more convenient user experience. In the IVC, Radio communications system may not guarantee the precise ranging and large capacity. Thanks to the VLC with a good directionality and ubiquitous LEDs, a HR-VLC will support a reliable and better ranging system for IVC. Furthermore, the utilization of SS helps reduce the interference of ambient light proportionally to the processing gain. In the HR-VLC in IVC system, PPM is used as a modulation scheme due to high average power efficiency. It is verified that the proposed system achieves better BER performance and more precise ranging.

Abstract: This paper introduces the empirical mode decomposition (EMD) method of the basic theory, problems and means to solve. Apply the approach to mechanical vibration signal containing a transient pulse processing and analysis carried out, and the wavelet time-frequency analysis methods are compared, the results show that it can effectively decompose nonlinear and non-stationary vibration signals, and has a self-adaptive, and in the time domain and frequency domain have better resolution capabilities, and the component with a more clear physical meaning. Due to its diversity of showing the results, you can make further precise analysis of a single component, and the transient signals can be effectively recognized, and can locate mutation point in time, describing the time-frequency localization properties. EMD, transient signals, mechanical vibration

Abstract: Image segmentation with low computational burden has been highly regarded as important goal for researchers. One of the popular image segmentation methods is normalized cut algorithm. But it is unfavorable for high resolution image segmentation because the amount of segmentation computation is very huge [1]. To solve this problem, we propose a novel approach for high resolution image segmentation based on the Normalized Cuts. The proposed method preprocesses an image by using the normalized cut algorithm to form segmented regions, and then use k-Means clustering on the regions. The experimental results verify that the proposed algorithm behaves an improved performance comparing to the normalized cut algorithm.

Abstract: This paper proposed a novel solution to track human face obscured largely in an image on the basis of Mean Shift Tracing Algorithm (MSTA). The improved approach aims to update the target model in real-time during the whole tracking process to avoid target losing. Local Binary Pattern (LBP) theory is chosen to improve the original MSTA here. The experimental result shows that our new solution has a better performance in target tracking under situations like face rotation and occlusion as well as in fast acquisition when faces reappear on the screen.

Abstract: Due to the interference of noise, filtering technology is applied to achieve gravity anomaly for airborne gravimetry. Kalman filtering and smoothing are discussed and implemented for data processing of airborne gravimetry in this paper. Firstly, the algorithms of Kalman filtering and smoothing are introduced. Then, the system model for solving the gravity anomaly is established which is based on the dynamic equation and the hardware design equations. Finally, the result of Kalman filtering and smoothing would be compared with digital FIR low pass filter, and it is proved that Kalman filter and smoother could obtain more accurate result than FIR low pass filter as that the solving error of Kalman filter and smoother is improved within 1 mGal compared with the theory standard obtained by GT-1A software.

Abstract: In order to construct exact soliton solutions of nonlinear evolution equations with variable coefficients. By using a transformation, the variable coefficient KdV equation with forced Term is reduced to nonlinear ordinary differential equation (NLODE), after that, a number of exact solitons solutions of variable coefficient KdV equation with forced Term are obtained by using the equation shorted in NLODE. As it showed above, this kind of method can be applied in solving a large number of nonlinear evolution equations.