Advanced Materials Research Vols. 753-755

Abstract: By the method of up-sub solutions, we consider a class of parabolic equations with nonlocal source. In the paper, we discuss the relation of the coefficients and the importance of the initial value. We get the sufficient conditions for the global existence and finite blow-up of the solutions.

Abstract: The fixed point iteration method was proposed to solve the forward kinematics of parallel kinematics mechanisms as a new method. In every iteration through the reverse kinematics the previous iteration intermediate position and posture determined stick orientation vectors. And the stick ports, which were obtained by multiplying stick orientation vectors and given stick lengths, determined the next iteration intermediate position and posture. The final position and posture satisfying the convergence conditions would be obtained. 6-SPS was taken as an example to verify that this method was available and correct. Compared with the Newton iteration, this method possesses more advantages, such as simpler formula, higher speed of the convergence, smaller program workload, and less influence of initial values on the convergence.

Abstract: Container loading problem with multi-constraints is usually modeled as a three-dimensional packing problem which is known to be NP-complete. This paper presents a novel approach to resolve the container loading problem via multi-objective optimization algorithm (NSGA-II). Specific encoding method and genetic operators are designed based on the mathematics description of the problem. The process also takes several practical constraints into account. Simulation results show that this approach is feasible and effective.

Abstract: Assigning the value of interest to each node in the network, we give a scale-free network model. The value of interest is related to the fitness and the degree of the node. Experimental results show that the interest model not only has the characteristics of the BA scale-free model but also has the characteristics of fitness model, and the network has a power-law distribution property.

Abstract: It is necessary to exactly classify the evaluating indicator of colliery security management, and the classification is also the foundation and prerequisite of colliery security management and evaluation. In this paper, reasonable classification to colliery security management and evaluation is carried by the method of grey incidence cluster. Grey incidence cluster is one kind of grey cluster evaluation in grey system theory. On the basis of lots of investigation, analysis and summary, sample nine coal mines, 13 safety management evaluation，calculate their gray absolute degree, taken given threshold r=0.75，classify these indicators appropriately to simplify the evaluation and examination indicator. Evaluating indicator cluster method is simple and convenient and has the character of strong applicability.

Abstract: A novel algorithm for face recognition is proposed. Firstly, illumination invariant feature of face image is exacted by the Nonsubsampled Contourlet Transform (NSCT) and small scale feature is exacted by the total variation model (LTV), respectively. Then the two type of features are fused into a new feature for face classification. The experimental results show that the proposed method can get very good effect in face recognition.

Abstract: Photograph method is a new scanning technique that provides cost-efficient and accurate computer system for log centering. A monochrome video camera is used to document images of a log in a perimeters rotation time. 3D digital contour of the log is constructed with the images. The biggest internal cylinder is calculated in software modules. Experiments in present study proved that photograph method was quite feasible in optimizing log centering with higher accuracy. The cost-efficient near-infrared images used in the devices were able to avoid the impact of changing illumination. This centering system can be utilized in small and medium enterprises.

Abstract: The restore algorithm of the image blurred by motion is proposed, and a mathematical model based on motion blur system is eomtrueted£®The Point spread function of the motion blur is given£®According to the characteristics of blurred images£¬the parameters of point spread function are estimated ,and three methods are introduced for image restoration. The three methods are inverse filtering of image restoration,Lucy-Richardson image restoration and Wiener image restoration.The principles of the three image restoration methods are analyzed. The motion blurred image restoration experiment is made. The results show that the visibility of the image is improved ,and the image restoration is more stable.

Abstract: For the poor performance of conventional filtering algorithms in removing salt and pepper noise from digital images under high noise density,an adaptive switching median filter algorithm based on BP neural network optimized by genetic algorithm (GA) is proposed to detect and remove salt and pepper noise from images. Firstly,the initial weights and thresholds of BP neural network are optimized by genetic algorithm.Then image pixels are devided into either signal or noise points by the trained network automatically. The detected noise points will be removed by adaptive switching median filter algorithm,but nothing to do with the signal points. Experiment results show that the proposed algorithm significantly outperforms the others and efficiently removes salt and pepper noise from digital images without distorting image details and textures .

Abstract: Iris recognition plays an important role in personal identification. In this study, we utilized CREASEG experimental platform to analyze the performance of some state-of-the-art image segmentation algorithms based on level set. Performance evaluation criteria include segmentation accuracy and computation time of pupil and iris localization. Four iris images were taken as experimental samples. The experimental results on those image samples demonstrate that Chan-Vese model achieve the best performance among all six algorithms. Furthermore, experimental results also show that energy functions play an important role, which should not make evolution curve to terminate at local minima or pass through the boundary. This study can provide certain referential significance in how to select image segmentation algorithm based on level set.