Applied Mechanics and Materials Vols. 411-414

Abstract: This paper argues that the current productive and multidisciplinary study in new literacy/literacies in digital era has lead to a complex situation full of kaleidoscopic terms and concepts, thus is in need of further exploration. Based on a critical ontology-based modeling and 5W1H maxim, the author proposes a 5-Ps framework to deconstruct and re-examine new literacy/literacies from five perspectives: Platform, Participant, Potential, Position, and Praxis. It is expected that this framework can help people to analyze and understand new literacies in study of electronics, computer, information technology as well as social sciences.

Abstract: In order to solve the problem that there isn’t an effective way to detect the violent video in the network, a new method using MPEG-7 audio and visual features to detect violent video was put forward. In feature extraction, the new method targeted chosen the features about audio, color, space, time, motion. Parts of MPEG-7 descriptors were added and improved: instantaneous feature of audio was added, motion intensity descriptor was customized, and a new method to extract dominant color of video was proposed. BP neural network optimized by GA was used to fuse the features. Experiment shows that these selected features are representative, discriminative and can reduce the data redundancy. Fusion model of neural network is more robust. And the method of fusing audio and visual features improves the recall and precision of video detecting.

Abstract: Putting forward a kind of star identification algorithm based on SOFM neural network. Firstly, using dynamic threshold selection algorithm what is based on supporting vector machine select guide stars to composite the navigation star database, then a recognition system what contain multiple parallel SOFM subnets for star map recognition is designed. Simulated recognition results show that the SOFM network can extract complex feature recognition navigation star from the chart. Compared with the traditional triangle algorithm, this algorithm has better recognition accuracy, robustness and real-time.

Abstract: In this paper, the embedded license plate recognition system based on TMS320DM642 is researched. During the design, median filter, threshold, and morphology closing operations are used to obtain license plate region, then segmented into disjoint characters for the character recognition phase, where the template matching is used to identify the characters. Embedded License Plate Recognition System, being smaller, has less power consumption with respect to software based LPR systems. The resulting hardware is suitable for applications where cost, compactness, and efficiency are system design constraints.

Abstract: Color medical images are widely used in the field of medical diagnosis. Image enhancement is one of the most important pretreatment methods which can enhance the quality of images. In this paper, a novel color image enhancement method using Y-H model and wavelet homomorhpic filtering is put forward. The chromaticity numbers matrix and intensity numbers matrix of color images are get using Young-Helmholtz (YH) transform. The chromaticity numbers matrix remains unchanged. Wavelet homomorphic filtering method is used to process intensity numbers matrix . The enhanced intensity numbers matrix and formerly chromaticity numbers matrix are processed by Y-H inverse transformation and disply in RGB color space. The method put forward in the paper is successfully used in color medical image enhancement. Experimental results show that the method have characteristics of nondistortion, better visual effect.

Abstract: There are more acquisition cells and long stop time on each cell because of the features of the new GNSS signal systems. So the average acquisition time of the new signals increase much longer than the former one. The new signal features was analyzed at first in this paper, and then the acquisition algorithm of QPSK and MBOC modulated signals was introduced separately. A Multiplexing Channel of Time-Division technique and Multi-frequency aid algorithm was proposed to increase the acquisition speed. This new fast acquisition algorithm significantly reduced the average acquisition time of the new GNSS signal systems.

Abstract: To get accurate frequency information from doppler signal is very difficult. In order to improve accuracy of doppler radio fuze, a design scheme which is not only using doppler signal amplitude information but also using doppler frequency information, and comprehensive and these two aspects of information give out the fuze work distance. The solution we give out in this paper is that we realize the STFT by application of parallel process mass on FPGA and together with quick frequency selection mechanism. Then we could get the frequency information from the doppler signal. Test results show that by using short time window Fourier transform, we could realize time-frequency analysis of doppler signal in allow system time delay.

Abstract: A large amount of processing data was accumulated within plant processing. And its necessary to use this data for plant processing as well as its administration. In this article the data mining technology and its utilization were discussed, according to research results, the fitting relationship is:Cu recovery (%)= -1.1221*lime dosage (Kg/t)+92.6, the lime dosage alteration effect on copper recovery are 1.41% absolutely and 1.66% relatively. The fitting relationship of copper concentrate grade and lime dosage is:Cu grade (%)= 0.0554*lime dosage (Kg/t)+19.271, the lime dosage alteration effect on copper recovery are 0.070% absolutely and 0.36% relatively. It can be concluded that the lime dosage has a great effect on copper recovery, and lime dosage is relative to the total metal minerals in the ore, because the fitting relationship of the lime dosage and metal minerals summation in the ore is:Dosage (Kg/t)= 0.0487*total metal minerals (%)+2.6441, the lime dosage show a positive relationship with total metal minerals in the ore.

Abstract: To overcome the uncertainty of the users neighborhoods in the recommendation algorithm of nearest neighbor, an improved collaborative filtering algorithm based on user clustering is proposed. This improved algorithm filters the users by their features, and the improved cosine similarity algorithm is used for the item similarity computation. Experiments on the MovieLens dataset showed that, compared with Lis collaborative filtering algorithm, the recommendation quality of the improved algorithm is more accurate and the category coverage is larger.

Abstract: In this paper, shuffled frog leaping algorithm (SFLA) is used to reduce the computation load of weighted subspace fitting (WSF) method for DOA estimation. As a recently proposed meta-heuristic algorithm, SFLA is suitable to solve the nonlinear multimodal optimization problem which WSF method encounters, so it offers an excellent alternative to the conventional methods in WSF-DOA estimation. Thus, the feasibility of SFLA applying to WSF-DOA estimation is analyzed, and its performance is compared with other popular meta-heuristic methods. Simulation results demonstrate that the proposed method is more efficient in computation and statistical performance.