Applied Mechanics and Materials Vols. 347-350

Abstract: in the 2-d numerical simulation of flash flood disaster, due to flood often occurred in the steep terrain and water flow rapidly changed, lead to that the calculated value is unstable and even the calculation diverge in the simulation. This paper presents a grid outflow correction method, which is based on the leap-frog finite difference format, through modifying the outflow rate of the grid circularly, to ensure the mass conservation in the whole process of computing. In the local dam bursting model, the simulated result comparison of the grid outflow correction method and the algorithm of implicit alternating direction on the mass conservation shows that, the new method can ensure the simulation accuracy and the numerical stability under the condition of steep terrain and moving boundary. According to the proposed method, the simulation analysis in the process of extreme flash flood disasters which happened in 2010 Zhouqu county in Gansu province was carried out. The comparison of simulation results and remote sensing estimation results shows that the deviation of the flood evolution time, speed and impact height are within 5%, and the consistency of evolution path is good, which verifies the validity of the algorithm.

Abstract: Due to effects of the light by water and other particles, the quality of underwater image will degrade. The traditional underwater image segmentation methods based on intensity and spectrum have difficulty in determining boundary. Inspired by the visual system of mantis shrimps, this paper constructed a feedback neural network model, in which the parameters were optimized using machine learning method. Based on this model, we combine the polarization and intensity information to achieve the underwater polarization image segmentation. The results of experiment prove that the neural network model designed in this paper can improve the accuracy of underwater image segmentation.

Abstract: Under the premise of in-depth understanding of Lucene full-text retrieval technology, this paper will apply it to the Mongolian text search. First, several key issues are proposed which are need to be addressed in achieving the Mongolian text search technology, and give the corresponding solutions to achieve the Mongolian full-text retrieval in Lucene. Second, this paper provides a fast, accurate and comprehensive Mongolian information full-text search service, played a key role in promoting the development of the Mongolian search engine.

Abstract: TO solve the applying constraints problem of D-S evidence conflict in multi-information fusion and achieve the systematic identification of evidence conflict,we introduce the K-L information on the distance function to describe the characteristics of the conflict between the evidence in the paper, construct the distance matrix defined the level of conflict of independent evidence in the whole system. Simulating results show that: the effective conflict identification of K-L information distance can complete the application constraint of D-S theorys synthetic rule, obtaining the optimized convergence results for synthesis of normal conflicting evidences, isolating the highly conflicting evidence.

Abstract: Scientific researches are always accompanied with a large number of complex data. Thus, it is important to view and analysis these data or objects in a much more efficient way. Viewpoint selection can improve both the speed of rendering and the efficiency of data understanding, became a hot topic in research, and is applied to various research areas. Viewpoint selection considerably influences the amount of information contained in the projected images, has been extensively studied in many areas such as: volume rendering, image-based modeling, object recognition, etc. most of the studies utilizing the information theoretical framework. This paper reviews the relevant literatures on this topic by summarizing the development and application of the technology, and put forward some ideas for future development.

Abstract: This paper puts forward a new dynamic round-robin (DYRR) packet scheduling algorithm with high efficiency and good fairness. DYRR algorithm introduces dynamic round-robin concept, that is, the allowance given to each of the flows in a given round is not fixed, but is related with the number of bytes sent of this and other flows of the last round scheduling. The time complexity of the DYRR algorithm is O(1). Results from performance simulation analysis shows that DYRR algorithm can effectively smooth output burst, realize fair scheduling, and have a good time delay characteristic.

Abstract: To judge the loop-nonnegativity of a matrix over an idempotent semiring and compute the plus-closure of when it is loop-nonnegative, a Plus_Closure_of_Matrix algorithm of complexity is constructed and proved. As a generalization of Floyd algorithm, Warshall algorithm as well as Gauβ-Jordan Elimination algorithm on idempotent semirings, this algorithm can also be used to solve some Algebraic Path Problems, Shortest Path Problems and the transitive closures of matrices over idempotent semirings even if the idempotent semirings have no completeness and closeness.

Abstract: With the rapid development of the network and the multimedia technical , how to protect the security of the multimedia informations becomes the popular topic on studies. As a new technique used to protect the copyright of digital productions , the digital watermark technique has drawn extensive attention . A digital watermarking algorithm based on discrete wavelet transform (DWT) was presented according to human visual properties in the paper . Then some attack analyses were given. Experimental results show that the watermarking scheme proposed in this paper is invisible and robust to cropping, and also has good robustness to compression, filtering, and noise adding.

Abstract: in order to further improve the retrieval accuracy of the image retrieval system based on shape feature. A partitioning image retrieval method based on regional division and polymerization is proposed in this paper. Firstly, an image is segmented by the regional division and polymerization method. Secondly, the shape and spatial features of different objects in the image are extracted by the invariant moment. Finally, images are retrieved by calculating the similarity of images, and five different types of images are tested by group. The experimental results show that it is more accurate for the algorithm to retrieve the users interested images.

Abstract: The rapid development of mobile technology has brought the change of learning. Mobile learning refers to a new kind of learning for the use of wireless communication technologies and mobile equipments to obtain the educational information, resources, and services. The framework of mobile learning system using wireless communication technology is proposed, a mobile learning support system with multiple wireless terminals is implemented which customs communication protocol to communicate with each other and uses multi-threading technology. The server push learning content to the terminal, learners receive information or to send information to the server through wireless terminal, to feedback information and interaction, and server processes terminal information, data analysis, and then sent to the learner. The system can be applied to a variety of mobile learning occasions, the application shows that the effect of the system effectively reduce the learning environment, and to increase the participation and interest of the learner, improve learning efficiency and learning outcomes.