Search

Face-detection methods based on cascade architecture have demonstrated a fast and robust performance. In most of these methods, each node of the cascade employs the simple Haar-like features from the central eye-nose-mouth region using the boosting method. However, it can be empirically observed that, in the deeper nodes of the boosting process, the non-face examples collected by bootstrapping are in fact very similar to the face examples, and the error rate of those feature-based weak classifiers is very close to 50%. Consequently, the performance of the face detector is hardly further improved. In this paper, we propose a novel and simple solution to this problem by imitating the characteristics of the human visual system. The main idea of our solution is to boost the cascade based on a hierarchical strategy, which employs the information from the central and surrounding parts of the face regions step by step. We argue that the context information about a face can be advantageously used in the deeper nodes of the boosting process when the features derived from the central region of the face do not provide any further benefit. Furthermore, we also propose a simplified Gabor feature to extend the feature set for the training of deeper nodes. Experiments show that our proposed method can improve not only the detection performance, but also the detection speed, by about 10% when compared to the original AdaBoost face-detection method for our implementation.

The synchronization of two 3-scroll hyperchaotic attractors is realized based on wavelet transform and single variables' feedback In the transmitter, one signal is decomposed by wavelet transform and the detailed information is removed, then the component with low frequency is reconstructed and sent into the channel In the receiver, the received signal is used as the feedback signal to realize the synchronization of two chaotic systems. Using this synchronous method, the transmitting signal is transported in compressible way, the system resource is saved, furthermore, because the transported signal is not a whole chaotic signal, the performance of security of the system is improved.

Recently least square estimator of L2 norm minimization and L1 norm minimization estimator are two popular algorithms in super resolution reconstruction. Thus in this paper, the pros and cons of L1 norm and L2 norm estimator are first analyzed, then they are weighted and combined, and adopting an approximate total variation regularization method [1][2], we proposed a weighted super resolution reconstruction algorithm based on an adaptive regularization parameter. The adaptive method regards the regularization parameter as a function of restored image. Experiments demonstrate that this method not only has better image edge-preserving and efficiently removes the visual artifacts and noise, but also enhances the quality of the restoration images and has better super resolution performance.

The weaknesses between L1 norm minimization estimator and L2 norm minimization estimator in the traditional super resolution reconstruction problem are analyzed. In this paper, L1 norm and L2 norm are weighted and combined to measure the data fidelity term, and based on an approximate total variation regularization method [1][2], a robust weighted and combined super resolution reconstruction algorithm is proposed. Simulation experiments show that this method not only enhances the quality of the restoration images and has better edge-preserving capability, but also efficiently removes the visual artifacts and is robust to noise.

For the dynamic topology and wireless channel features of mobile ad hoc networks, its security problem is always being a research focus. In the aspect of security routing protocol, the problem of how to enhance the security of mobile ad hoc networks was discussed in the paper. Based on the Ad Hoc networks structure and routing characters analysis, the requirements of Ad Hoc networks security routing protocol was brought forward. Then the insecure factors of current network routing protocol were analyzed in detail. To solve these insecure problems, corresponding solutions and material examples were proposed, meanwhile, an advanced strategy to optimizes the security route reconstruction was proposed.

The fast, efficient and reliable dynamic routing protocol is the basis to ensure highly reliable work of the Mobile Ad hoc Network (MANET). For the shortcomings, the traditional proactive, reactive and hybrid routing protocol can not meet the requirements of reliable work of MANET. To address the problem and combined with the application requirements of small-scale MANET, a dynamic adjacent matrix routing protocol (DAMRP) that using adjacent matrix to describe network dynamic topology was brought out in the paper. The research and practice indicates that DAMRP overcome the shortcomings of traditional routing protocols and reduce the overhead, which is an appropriate efficient routing protocol for small scale MANET.

In this paper, a novel design for improving the fractional order differential filter is put forward. By analyzing the frequency characteristic of typical fractional order differential filter, it can be seen that these kinds of differential filters have merits and demerits respectively and also could be complementary each other. So based on these features, three kinds of novel first order differential filters are constructed by the interpolated method. And then we choose a differential filter from these three kinds of filters which has much better frequency characteristic, also the improved IIR-type fractional order differential filter will be obtained by the method of continuous fraction expansion (CFE). The experiment result shows that the frequency response of the improved fractional order differential filter is more approximate to the ideal fractional order differential filter. And it also shows that the method put forward in the paper can improve the performance of the fractional order differential filter obviously under the premise of not increasing the structure complexity of the filter.

Variety of facial pose and expression will affect result of face recognition. Traditional methods analyze facial pose and expression through every pixel of image, on the contrary, the paper analyzes pose and expression through geometric relationship among images. The paper uses locally linear embedding (LLE) which computes low dimensional, neighborhood preserving embedding of high dimensional data. The algorithm not only finds the nonlinear structure of data effectively, but also is invariable to translation and revolution. The experimentation shows that the algorithm can reflect slight variety of facial pose and expression, and indicates that low dimensional distribution of various pose and expression is identical to different persons and original image can be reconstructed exactly from neighboring images.

The paper puts forward and discusses a 1/2 order net-grid-type analog fractance circuit. Furthermore, it finds out that our 1/2 order net-grid analog fractance circuit is superior to the classical 1/2 order tree-type one by comparing the structure and function of each other. On this bases, it further puts forward and discusses a general recursive net-grid-type analog fractance circuit for any order fractional calculus. The results educing by the paper are the basis for further theoretic research and engineering implement for structuring fractance circuit of any order fractional calculus

This paper addresses the problem of noise reduction under stationary and non-stationary environments, which based on the wavelet transform. This algorithm can overcome the deficiency of the conventional algorithms of noise reduction, which were only efficient for stationary environments and have large level of signal residual noise. The algorithm is based on the different amplitude value change of signal and noise and their distributing character in the wavelet domain, by this way we can find the site and the value of the noise in the wavelet domain. Experiments confirm that the noise reduction by proposed algorithm is effective to reduce the noise under stationary and non-stationary environments.

A face detection method in color images using multi-resolution sub-images fusion combined with support vector machine (SVM) is proposed. Firstly the skin images are obtained by two different color modes (HSI and YCbCr). Secondly, multi-resolution sub-image is gained by wavelet analysis of skin area image. Thirdly fusion image can be obtained by sub-image LH and HL according to pixel fusion criterion. Fusion images should be incorporated if necessary, candidate face area can be detected in fusion image. Finally face images can be validated by using SVM. Experiment result shows that, detection rate of faces in a complex background image achieves 95.8%. This method is not effected by gesture, expression and rotation but the distance of the faces in a group photo.

Fractional calculus is a novel powerful tool for non-linear signal processing. This paper realizes the arbitrary order fractance of fractional calculus based on one order Newton process by solving the positive real root of the n-order equation as the approximation of the fractance. The condition under which the one order Newton process can be convergent is discussed. For semi-order fractance realization, the Steffensen method is used to accelerate its iteration process, which shows a good performance. We compare our method with others and present the analog passive realizations scheme of the arbitrary order fractance in the end.

In the first, the paper analyzes the classical tree type analog fractance circuit for frac12 order fractional calculus. In the second, it puts forward and discusses our three new types analog fractance circuits for frac12 order fractional calculus. It compares the structure and function of our three new type's fractance with the classical one respectively. It is true that the frac12 order net-grid type analog fractance circuit is the best one of the four types. In the last, it studies some theoretical and engineering problems that need to be deeply researched in the area of the fractional calculus applications to electronic communication and information processing

In the first, the paper puts forward a frac12 order H type analog fractance circuit. In the second, it further puts forward any fractional order H type analog fractance circuit for any order fractional calculus. In the last, it is proved to performance correctly and efficiently by computer simulation and circuit analog. The result educing by the paper is the basis for further theoretic research and engineering implement to structure fractance circuit for any order fractional calculus

The abstract is to be in fully-justified italicized text, at the top of the left-hand column as it is here, below the author information. Use the word "Abstract" as the title, in 12-point times, boldface type, centered relative to the column, initially capitalized. The abstract is to be in 10-point, single-spaced type, and up to 150 words in length. Leave two blank lines after the abstract, then begin the main text

This paper proposes a new way for ultrasound color flow velocity estimation and display. This method sets up a mapping table which directly relates to the dynamic range of the velocity display by using an inverse mapping instead of calculating the arctangent value directly. This method is ideal for software implementation and offers an interactive way to user for changing the dynamic range of flow velocity to increase the display resolution.