Recently, active contour models based on local information have emerged in image segmentation. These models are more robust to local variations of region of interest. But it also brought some new problems, such as loc...
详细信息
Recently, active contour models based on local information have emerged in image segmentation. These models are more robust to local variations of region of interest. But it also brought some new problems, such as local minimum, higher computational cost. To effectively alleviate these problems, this paper presents a novel fast active contour model driven by global-local statistical energy. Firstly, a new local statistical model is constructed to capture the border of object more accurately. Secondly, global image features of image are integrated with local features so as to create an improved model which can avoid local minimum effectively. In addition, an accelerating factor is introduced into global-local statistical model in order to drive contour evolution rapidly. Finally, we carry out experiments on synthetic and real data, and results compared to relevant models have demonstrated robustness, accuracy, and time efficiency of our proposed model.
Although many kinds of clustering algorithms are proposed, there has been much less work on the incremental clustering. Inspired by the artificial immune systems, the authors apply it to the incremental clustering, pr...
详细信息
Although many kinds of clustering algorithms are proposed, there has been much less work on the incremental clustering. Inspired by the artificial immune systems, the authors apply it to the incremental clustering, propose a novel incremental clustering algorithm called ICAIS. The algorithm mainly uses the mechanism of immune response of the adaptive immune system. The primary immune response corresponds to forming the new clusters. The secondary immune response corresponds to recognizing the data belonging to the existed clusters. The experiments show that ICAIS has advantages in recognizing new pattern, the effect is better.
A novel dynamic software watermark scheme based on the Shamir threshold and branch structure is presented. First, we split the watermark into a set of shares using the Shamir threshold scheme. Second, these values are...
详细信息
A novel dynamic software watermark scheme based on the Shamir threshold and branch structure is presented. First, we split the watermark into a set of shares using the Shamir threshold scheme. Second, these values are encrypted with the DES block cipher that forms the watermark shares to be embedded into different methods of program according to the dynamic behavior of the branch structure. Our scheme can withstand most semantics-preserving attacks and can retrieve the original watermark based on partial information. Simulation tests show that our scheme is very robust, stealthy and has a high price performance rate compared with other methods.
The nonlinearity in normal and epileptic electroencephalogram (EEG) signals is investigated in this paper by the delay vector variance (DVV) method,which determines the degree of nonlinearity of the tested time series...
详细信息
The nonlinearity in normal and epileptic electroencephalogram (EEG) signals is investigated in this paper by the delay vector variance (DVV) method,which determines the degree of nonlinearity of the tested time series by comparing the target variances of the tested time series to those of the corresponding surrogate time *** results of numerical experiments show that both normal and epileptic EEG signals are of nonlinearity,whereas epileptic EEG signals are of higher degree of nonlinearity than normal EEG *** on this,it is proposed that degree of nonlinearity could provide useful information for epileptic seizure ***,the degree of nonlinearity of epileptic EEG time series fluctuates more briskly than that of normal EEG time series.
After analyzing the specific security requirements of the wireless communication network based on IEEE802.16, a series of key issues about security methods and techniques were researched. A new authentication protocol...
详细信息
After analyzing the specific security requirements of the wireless communication network based on IEEE802.16, a series of key issues about security methods and techniques were researched. A new authentication protocol for wireless mobile environment was proposed and designed. Then, its security was analyzed from the angle of technical realization and formal analysis to verify whether the original safety goals were achieved. Especially, formal analysis was mostly done. Finally, system testing was also carried out and the results were analyzed to verify function and performance of the new protocol.
A novel time delay neural network is proposed for dynamical system control. In this work, A continuous recurrent neural network with time delay neurons in hidden layer is constructed, and the novel training algorithm ...
详细信息
A novel time delay neural network is proposed for dynamical system control. In this work, A continuous recurrent neural network with time delay neurons in hidden layer is constructed, and the novel training algorithm and control law independent of delay are developed based on Lyapunovpsilas stability approach. Using the proposed method, the control error converges to a range near the zero point and remains within the domain throughout the course of the execution. The usefulness and validity of the presented algorithm are examined by numerical experiments.
Association rule discovery is one of kernel tasks of data mining. Concept lattice, induced from a binary relation between objects and features, is a very useful formal analysis tool. It represents the unification of c...
详细信息
Association rule discovery is one of kernel tasks of data mining. Concept lattice, induced from a binary relation between objects and features, is a very useful formal analysis tool. It represents the unification of concept intension and extension. It reflects the association between objects and features, and the relationship of generalization and specialization among concepts. There is a one-to-one correspondence between concept intensions and closed frequent itemsets. This paper presents an efficient algorithm for mining association rules based concept lattice called Arca (Association Rule based Concept lAttice). Arca algorithm uses concept-matrix to build a part of concept lattice, in which the intension of every concept be put into one-to-one correspondence with a closed frequent itemset. Then all association rules are discovered by 4 operators which are defined in this paper performed on these concepts.
The necessary and sufficient condition for the existence of prewavelets with finite decomposition and finite reconstruction is presented. Especially when r = 1, we show that {φj(x-k) | 1 &le j &le r, k ∈ s} ...
详细信息
The necessary and sufficient condition for the existence of prewavelets with finite decomposition and finite reconstruction is presented. Especially when r = 1, we show that {φj(x-k) | 1 &le j &le r, k ∈ s} is shift-orthogonal if there exist prewavelets with finite decomposition and finite reconstruction, i.e., non-trivial prewavelets with finite decomposition and finite reconstruction does not exist. As an example, when r = 2 we construct a prewavelet with finite decomposition and finite reconstruction where the scale function is not shift-orthogonal, which demonstrate that there exist vector prewavelets with finite decomposition and finite reconstruction.
A new iris identification algorithm with tree-structured wavelet transform and dominant frequency channel energy is represented in this paper . After the iris image is preprocessed firstly, the normalized iris texture...
详细信息
A new iris identification algorithm with tree- structured wavelet transform and dominant frequency channel energy is represented in this paper . After the iris image is preprocessed firstly, the normalized iris textur...
详细信息
A new iris identification algorithm with tree- structured wavelet transform and dominant frequency channel energy is represented in this paper . After the iris image is preprocessed firstly, the normalized iris texture image is obtained and the wavelet is decomposed on selected base. Secondly the decomposed path is determined through some restriction condition. And the selected energy channel and channel value are regarded as iris feature. Finally, the experiment result verifies the correctness and validity of the approach.
暂无评论