Advanced Materials Research
Vol. 937
Vol. 937
Advanced Materials Research
Vol. 936
Vol. 936
Advanced Materials Research
Vol. 935
Vol. 935
Advanced Materials Research
Vol. 934
Vol. 934
Advanced Materials Research
Vol. 933
Vol. 933
Advanced Materials Research
Vols. 931-932
Vols. 931-932
Advanced Materials Research
Vols. 926-930
Vols. 926-930
Advanced Materials Research
Vol. 925
Vol. 925
Advanced Materials Research
Vol. 924
Vol. 924
Advanced Materials Research
Vol. 923
Vol. 923
Advanced Materials Research
Vol. 922
Vol. 922
Advanced Materials Research
Vols. 919-921
Vols. 919-921
Advanced Materials Research
Vol. 918
Vol. 918
Advanced Materials Research Vols. 926-930
Paper Title Page
Abstract: Owing to the reality of different focus, temperature localization, etc, there usually generate all kinds of image of local fuzzy, In solving such problems, the fusion technology of grayscale image has have already developed into a rather mature stage, however, in the color images field it still exists some shortcomings and need to be improved gradually. In order to dispose the problem that the method of traditional wavelet transform fusion is used in color image fusion will lead to the color difference, this paper proposes the method strategy based on the wavelet transform and R, G, B three color channel binding type fusion, it can achieves the effect of fuzzy image without producing the Color difference.
2902
Abstract: Of all the active learning research, the study on the active learning algorithm for SVM is much less. In this paper, based on K-Nearest Neighbors (KNN), we propose a new SVM active learning algorithm. The algorithm first collects the potential informative samples to form a potential informative sample set, and then in this sample set, based on KNN it evaluates the sparseness for each sample. The sample that locates at a sparser region is taken as an informative one, and is selected for training. Experimental results show that the proposed algorithm can greatly improve the classification performance, and can efficiently accelerate the convergence of the classifier.
2906
Abstract: Image repair using the digital image processing technology has become a new hot point in the cultural relic protection. To study of ancient fresco restoration techniques, A novel algorithm of local statistic enhancement image is proposed in this paper for the reparation of ancient fresco. The modified amplified function and the rubber band conversion algorithm are used as the jugement criterion, and the optimal paremeters are searched by the genetic algorithm (GA). Experimental results show that the quality of images is improved compared with the traditional.
2910
Abstract: A semi-supervised regression algorithm based on co-training with the same type KNN required large amount of storage capacity and shown scarcely to improve the regression precision further after several iterations. This paper puts forward a kind of a semi-supervised regression algorithm based on co-training with SVR-KNN, which fully combined the advantages of SVR and KNN in a semi-supervised learning respectively, avoided the learning ability limitations of a single type of learning. Finally, comparative experiments of the semi-supervised regression algorithm based on co-training with SVR-KNN and the other two co-training algorithms with same type learners was conducted, the result shows that the algorithm proposed here works much better in improving regression precision and generalizing the regression model.
2914
Abstract: Topology aggregation is very important for multi-domain network. To deal with wavelength and timeslots constrains in multi-domain WDM-TDM networks, a topology aggregation scheme is proposed in this paper. The proposed scheme is based on asymmetric star with bypasses structure and comprehensively considers the information of multi-path between each border nodes pair. Results show that our proposed scheme obtains a better tradeoff between information accuracy and information reduction.
2919
Abstract: The accurate maintainability prediction and evaluation of software applications can improve the designing management for these applications, thus benefiting designing organizations. Therefore, there is considerable research interest in development and application of sophisticated techniques which can be used to build models for both predicting and evaluating software maintainability. In this paper, we investigate some ideas based on Machine Learning, Natural Language Processing, Fuzzy Logic, and Systematic Model of Software Maintenance. The idea to compute Interactive Index and the maintainability of software system is useful to study the relation between maintainability prediction and maintainability evaluation in the whole software process. An model basing on fuzzy matrix and BP neural network is built up. It’s approved that there are application value of using this model based on BP neural network to predict and evaluate the software maintainability.
2924
Abstract: CS used in MRI image reconstruction is a research hotspot recent year. For the problem that reconstruction rate slow in MRI image reconstruction based on CS .Type acceleration Bregman iterative regularization algorithm to solve the MRI imaging sparse model ,and use the accelerate gradient method and the Restring in processing . The simulation data express this algorithm effective enhance the reconstruction rate, It’s have positive meaning in MRI image reconstruction that have strict in time requirement.
2928
Abstract: In order to quickly and accurately find the community structure of complex networks ,This article start from the similarity of the node ,Proposed a new community discovery algorithm. Introduced similar values and custom node value Q during the process of algorithm design ,Firstly , To Select the nodes with the largest similarity value by calculating the similarity between nodes ,Then to decide to join and expand the nodes by calculating the Q value is greater than 0 or not. Repeat the above process, you can get the whole network of community structure, The process does not require any auxiliary information or other seed nodes. Applied to the actual network experiment results verify the feasibility of the algorithm.
2932
Abstract: Optical flow is an important kind of video motion tracking algorithm, and Lucas-Kanade (LK) algorithm is an effective differential method in terms of calculating optical flow. The 3D Gaussian smoothing filter is properly introduced in the image preprocessing stage of the LK algorithm, which makes it possible to increase the correlation of the adjacent pixels in the time axis, improve the blur effect of the video image and overcome the 2D Gaussian filters disadvantage that is not suitable for the video image processing. More importantly, the optimized 3D non-Gaussian matching filter is chosen during the 3D derivative calculating, and it is capable of reducing the error rate of the velocity vector calculation and enhancing the calculation accuracy of the optical flow.
2938
Abstract: The paper proposes an algorithm based on the self-organizing Kohonen’s SOM to resolve the difficulties brought by the information fusion in the color image segmentation. First, considering the relationship of NBS distance and human perception, the image’s information is transformed from the RGB to the Munsell color space. Combining the spatial information, the initial segmented regions are formed by the kohonen’s SOM training according to the computational method of distance provided in the paper. Second, the initial regions are merged until the termination rule of the merging process is contented. The algorithm syncretizes the color and spatial information, which is demonstrated that segmentation results hold favorable consistency in term of human perception consistency by a great deal of experiments.
2942