Advanced Materials Research Vols. 706-708

Abstract: Face detection is the first step of face recognition, and is a very active research topic in the filed of computer vision and pattern recognition. A skin color model based face detection method for chromatic images is proposed in this paper. The H-CgCr skin color model is established by taking advantage of the color pixels clustering distribution in the HSV and YCgCr color space. The noises are eliminated based on skin color segmentation, and the face candidate region is judged according to knowledge-based, finally, the position of the face area is marked by the box. The experimental results demonstrate that the proposed method is feasible and effective.

Abstract: This paper introduces an effective face detection method based on log-Gabor filters. First, Information contained in the log-Gabor transformations is analyzed. Next, 5-scale and 8-orientation log-Gabor filters are constructed to extract feature vectors. Then, a three-layer feedforward network is created for feature classification. After training the network, we test 80 samples. From the experimental result, log-Gabor filters based method has comparable detection performance with Gabor filters based method. Therefore, log-Gabor filters can encode the images more efficiently.

Abstract: A new method for tracking of moving human objects is presented. The developed algorithm is based on B-spline Snake model. The Snake algorithm accesses the object contour through minimizing the energy; the object contour in each frame is presented through three B-spline curve. This approach gets the change image with double thresholds image segmentation between the neighboring frames, it could detect moving human objects with high quality and locate objects approximately. The contour extracts from the last frame put at the approximate position as the B-spline Snake model initialization. It accomplishes the tracking of moving human objects.

Abstract: Matrix ordering is a key technique when applying Cholesky factorization method to solving sparse symmetric positive definite system Ax = b. Much effort has been devoted to the development of powerful heuristic ordering algorithms. This paper implements a sparse matrix ordering scheme based on hypergraph partitioning. The novel nested dissection ordering scheme achieve the vertex separator by hypergraph partitioning. Experimental results show that the novel scheme produces results that are substantially better than METIS.

Abstract: In order to evaluate the ability of simulation of the software MapleSim, this paper introduces MapleSim software and its features. Experiments on two independent domains and one multi—domain physical are modeled and simulated. The multi—domain physical system has several outputs that can be shown as visible charts. Experiments demonstrate the MapleSim ability of modeling and simulation under complex condition. The result of simulation has highly reliability.

Abstract: Research on remote fault diagnosis of wind power has great significance of protecting grid security and secure operation of wind farm This paper presents the remote fault diagnosis system of wind power based on cloud computing and thoroughly analyzes construction demands and functional structure. Experimental results show that this system can effectively analyze mass monitored data and accurately assess equipment failure. This provides a new idea for cloud computing applied to wind farm fault diagnosis.

Abstract: Giving attention to the benefits of the passengers and agency, this paper adopts the true value of the coding method using the start time as the variable and uses the penalty function method to add a variety of constraints to the objective function when constructing the fitness function,which simplifies the calculation. Finally, the simulation results are obtained by using the improved Genetic Algorithm for solving the non-uniform grid schedule. Results show that the improved Genetic Algorithm can find the approximate best result in the huge search space of optimization, and greatly increased the computational efficiency.

Abstract: Simulation technology was a very important technology in the parallel signal processing research. A type of antomatic controlling simulation system was introduced, which was built by taking LabVIEW as the software platform, using LabVIEW user interface interacting with the Simulink model. Simulation of processing often requires algorithm and performance evaluation of parallel environment. Signal processing often requires simultaneous acquisition of multi-channel analog signals. Taking dual-channel A/D conversion processes as an example, using LabVIEW with VC, it achieves a simulation of parallel signal process of A/D conversion process.

Abstract: Change impact analysis is very important to the object-oriented software development and maintenance. Aiming at the problem of change prediction at design level, dependency relationship among classes is analyzed firstly, and the relation of correlative class is described by conditioned probability. Then the prediction of software maintainability method base on probability is proposed. The proposed method can compare software maintainability of different design projects. Lastly, the feasibility and practicability of this method has been proved through one case.

Abstract: In the process of speaker recognition, specific algorithm is adopted to classify differentspeakers. In this paper, Multiclass Core Vector Machine (MCVM) is used to the solve speakerrecognition problem. At first, CVM transform quadratic programming of traditional SVM into theMinimum Enclosing Ball (MEB) problem, which significantly reduces the complexity ofcomputation and then, defining an SVM with vector valued output. At last, Experimental results showthat the algorithm is feasible and effective for speaker recognition.