Chen Deyun

A Novel Principal Component Analysis Neural Network Algorithm for Fingerprint Recognition in Online Examination System

To solve the authentication problem in online examination system for large-scale, a novel principal component analysis neural network algorithm for fingerprint recognition is presented. Based on the introduction of the basic principles of feature selection and feature extraction for principal component analysis¿Construction of Symmetric subspace model based on principal component analysis neural network, and the convergence of Symmetric subspace algorithm is analyzed.The feasibility of using this algorithm for fingerprint recognition problems is also discussed. Algorithm to meet the convergence conditions and to simplify the complex pre-processing steps, greatly reducing the computational complexity, improve the speed of the identification. Experimental results indicate that the algorithm can obtain a higher recognition rate compared with BP neural network recognition algorithm and this new algorithm presents a feasible and effective way to research on fingerprint recognition algorithm for the examination.

A novel multiple-electrodes excitation method for electrical capacitance tomography system

Aiming at the non-uniformity problem of sensitivity distribution caused by the “soft field” effect of the static electric field in the measurement region of capacitance sensor, on the basis of deeply analyzing the capacitance measuring principle in electrical capacitance tomography system, an triple electrodes excitation and double electrodes examine method is proposed. By using the new method, a more uniform electric field can be got, and the independent measurement data is also increased, which improves the ill-posed of the problems, enhances the uniformity of capacitance sensitivity. The electromagnetic properties of the sensor are simulated by utilizing the ANSYS and MATLAB software. The experiment and calculation results show that the difficulty of measurement circuit is reduced, the uniformity of capacitance sensitivity is improved greatly with triple electrodes excitation and double electrodes examine method compared with singe electrode excitation method, and the quality of the reconstructed image is also improved.

An image reconstruction algorithm based on artificial fish-swarm for electrical capacitance tomography system

According to the fundamental principles of electrical capacitance tomography (ECT), a new ECT algorithm optimized Radial Basis Function (RBF) neural network algorithm, which is based on Artificial Fish Swarm Algorithm (AFSA) is proposed against the “soft field” effects and ill-conditioning problems in ECT technology. After giving the mathematic model of the algorithm, this paper also applies the AFSA to the training process of neural networks to compare with the traditional neural network algorithm. At last, a conclusion that with little error, high quality and fast convergence rate, etc. The ECT image reconstruction algorithm which is based on AFSA and the optimized RBF neural networks providing a new way for the ECT image reconstruction algorithm is reached.

Encoded light image active feature matching approach in binocular stereo vision

Binocular stereo vision is one of the most effective ways to obtain the 3D information of the object, of which the stereo matching is the most important and difficult part. In this paper, we propose an active feature matching approach based on encoded feature patterns, which add active features to the object image through projecting Gray code light to the object. Besides, we set the constraints, propose the searching strategy and take the pixel region code value as the match unit to achieve rough matching. Focused on the problem that several pixels share the same region code value, we set the sliding window and implement precise matching according to the similarity measurement function. Furthermore, we establish the experimental system and carry out experiments on multiple objects. Experimental results show that the proposed approach is feasible and under the same condition, the measurement accuracy of the proposed approach is high than that of the Gray code structured light approach.

A Superresolution Image Reconstruction Algorithm Based on Landweber in Electrical Capacitance Tomography

According to the image reconstruction accuracy influenced by the “soft field” nature and ill-conditioned problems in electrical capacitance tomography, a superresolution image reconstruction algorithm based on Landweber is proposed in the paper, which is based on the working principle of the electrical capacitance tomography system. The method uses the algorithm which is derived by regularization of solutions derived and derives closed solution by fast Fourier transform of the convolution kernel. So, it ensures the certainty of the solution and improves the stability and quality of image reconstruction results. Simulation results show that the imaging precision and real-time imaging of the algorithm are better than Landweber algorithm, and this algorithm proposes a new method for the electrical capacitance tomography image reconstruction algorithm.

Research on key technology of Internet of Things based on FRID

The Internet of Things refers to a network of internet-enables objects, together with web services that interact with these objects which bases on appointed treaty to make information exchange and communication in order to achieve intellectual identification, location, tracking, monitoring and management. Since 2009, the the Internet of Things concept in the global and domestic has become hot flashes, which called the third wave in global Internet information industry after computer and internet. This paper analyzes the Internet of Things application perception technology, network communication technology, data encryption technology, network security technology, and other key technologies based on FRID technology in recent years which will surely promote all their important roles in the IOT.

Two-phase flow parameters measurement and Gauss-Newton image reconstruction algorithm for Electrical Capacitance Tomography

To solve the problem of two-phase flow measurement for electrode electrical capacitance tomography (ECT), on the basis of analyzing ECT system measurement principle, we constructed the mathematical model of two-phase flow distribution information according to double integral infinitesimal principle, and formed a set of 12-Electrode two-phase ECT system. And to speed up the slow convergence rate of ordinary iteration method, a Gauss-Newton algorithm is improved on the basis of analyzing mechanism of the algorithm, and the feasibility of using this algorithm for ECT problems is discussed. It shows that it is easy to meet the convergence condition and error of image reconstruction is small. Experimental results and simulation data indicate that the algorithm can provide high quality images and favorable stabilization compared with LBP, Landweber and conjugate gradient algorithms and this new algorithm presents a feasible and effective way to research on image reconstruction algorithm for Electrical Capacitance Tomography System.

A novel quasi-newton image reconstruction algorithm for electrical capacitance tomography system

To solve the electrical capacitance tomography(ECT) technology “soft field” effect and pathological problem, a Quasi-Newton new algorithm for electrical capacitance tomography is presented. On the basis of analyzing ECT system measurement principle, constructing corrector formula of secant approximation algorithm in second-order information items of objective functions. The feasibility of using this algorithm for ECT problems is also discussed. It shows that it is easy to meet the convergence condition and error of image reconstruction is small. Experimental results and simulation data indicate that the algorithm can provide high quality images and favorable stabilization compared with LBP and conjugate gradient algorithms and this new algorithm presents a feasible and effective way to research on image reconstruction algorithm for Electrical Capacitance Tomography System.

A spatially compact mixture model for image segmentation

Due to the coupling of model parameters, most mixture models based on Markov random field for image segmentation cannot be computed directly by EM algorithm. The paper introduces spatial correlation into mixture models by two ways: First, visual description and spatial position of pixels are considered as co-occurrence data and a joint Gaussian mixture models are built to make use of spatial constraint. Second, for the visual part of the mixture models, the pixels are sampled by the weights according to the visual similarity with their neighborhood. So the noise pixels rely more on the position and less on the visual description in the clustering procedure. Because the spatial distributions of parametric model cannot fit arbitrary shape of the segmented objects, the paper adopts an over-segmentation scheme and develops a fast EM algorithm. Experiments using synthetic and natural images are presented to show the effectiveness of the proposed algorithm.

New methods for the construction of Voronoi Diagram and the nearest neighbor query

The existing methods for the construction of Voronoi Diagram and the nearest neighbor query have several disadvantages. In view of the disadvantages, the new methods were studied in detail. To construct the Voronoi diagram effectively, the datasets were grouped in advance, and then the divide-and-conquer method combined with the increment method was used to generate Delaunay triangulation. By generating the Voronoi diagram from Delaunay triangulation, the Creat_Vor algorithm was proposed. In order to effectively query and update the nearest neighbors of the given points, the methods based on the spatial grids and the Voronoi diagram were studied. The VGride_NN algorithm and VGride_BNN algorithm were put forward. Theory and experiments show that the new methods have advantages in dealing with the dynamic dataset.