Applied Mechanics and Materials Vols. 543-547

Abstract: L2C is a new civilians signal launched by the modernized GPS Block IIR-M satellite. This paper studies L2C acquisition algorithms with the implementations on MATLAB. Circular correlation is utilized to implement the acquisition algorithm. The input satellite signal is collected by hardware front-end and the local code then simulated by software. The input data after frequency reduction processing and the local simulated code are converted into the frequency domain by means of FFT (Fast Fourier Transform). After performing circular correlation, the initial phase of the CM code is attained and the carrier frequency is found with the resolution of 50Hz.The effectiveness of the acquisition algorithm is finally verified through the actual satellite experiments.

Abstract: This paper studies a new QR decomposition adaptive filtering algorithm for acoustic echo cancellation (AEC). Based on the p-TA-QR-LS algorithm [ and an efficient voice activity detection technique, the proposed algorithm can distinguish the significant and insignificant input data periods. The resultant variable mode p-TA-QR-LS algorithm can work between two modes (p=1and N) and is thus suitable for AEC application where reusing significant input data can enhance convergence and the computation cost can be saved when the input is relatively weak.

Abstract: In order to extract the expression features of critically ill patients, and realize the computer intelligent nursing, an improved facial expression recognition method is proposed based on the of active appearance model, the support vector machine (SVM) for facial expression recognition is taken in research, and the face recognition model structure active appearance model is designed, and the attribute reduction algorithm of rough set affine transformation theory is introduced, and the invalid and redundant feature points are removed. The critically ill patient expressions are classified and recognized based on the support vector machine (SVM). The face image attitudes are adjusted, and the self-adaptive performance of facial expression recognition for the critical patient attitudes is improved. New method overcomes the effect of patient attitude to the recognition rate to a certain extent. The highest average recognition rate can be increased about 7%. The intelligent monitoring and nursing care of critically ill patients are realized with the computer vision effect. The nursing quality is enhanced, and it ensures the timely treatment of rescue.

Abstract: In order to realize rapid alphabet recognition, the paper proposes an alphabet recognition method based on computer vision optimization technical which can also extract the classification features. Experimental results show that the obtained variance value of the test image and the standard image obtained by the proposed method is the minimum which indicating the method can achieve correct match, effective classification, and provide a great method of identification.

Abstract: In traditional algorithms for virtual 3D image synthesis, the high complexity of virtual objects will cause serious image drift and reduce the fidelity of virtual 3D image synthesis. To solve the problem, this paper proposes a virtual 3D image synthesis method based on characteristics' dynamic optimization algorithm. According to edge data of virtual objects, the method calculates the dynamic change center to provide accurate data basis for virtual 3D image synthesis; extracts image characters of virtual images, establishes mesh model for image features, obtains the spatial location of the characteristic points to complete 3D image synthesis and improve the fidelity. Experimental results show that the algorithm is simple simulation method with strong practicality.

Abstract: Considering the problem of dynamic positioning Systems for the slowly-varying disturbances, a new kalman filter using position and acceleration feedback is presented. The kalman filiter separates the WF motions from the measured position and linear acceleration, and estimates the LF position, velocity and acceleration. The stability of this filiter is proven by applying input-to-state (ISS) stability theory. Finally, the computer simulation is given to demonstrate the effectiveness of the proposed method.

Abstract: For enhancing printing images quality so as to meet human visual needs, two ways of scanners clarity compensating by optical unsharp-mask and electronic second differential had been devoted based on investigating the cause of cut down on presswork definition produced by printing, the advantages of the twofold method which combine the two was also discussed in the thesis. The method of applying the technique was proposed in digital image processing. Tests had been done on standard images by the suggested method in Matlab, normal ways of USM and second differential were also used for comparing, printing proofs were made after manipulation for subjective evaluation on sharpening degree, statistic analysis on assessment data was also made, which shows that: the second differential method will bring the highest clarity on images with the side-effect of too heavy edges, USM can get relative mild margins and slightly lacking of distinct, the twofold method get preferable contour strengthen and smoothness, it can be used in printing image sharpening.

Abstract: In face recognition with single training sample cases the question of the low recognition rate, proposed a method of a kind of three layer virtual image. Using singular perturbation method to highlight the facial features, increase the quantity, scale change and sample attitude through the geometric transformation method, the spatial distribution of the improved method based on sample distribution. The experimental results on ORL face database show that this method can effectively on the single sample problem in the training sample pretreatment.

Abstract: This paper builds sample analysis data sets through the analysis of specific commands behavior in command behavior experiment, and selects typical commanders as reference sample, proposes a command pattern classification model based on k-means clustering algorithms. The command model calculation is simple, easy for computer implementation, and is of great significance in understanding and mastering specific commanders characteristics, and application to practical operations and training.

Abstract: Nonlinearity of power amplifier always results in spectrum spreading on input signal. When using digital predistortion technology to compensate power amplifier's nonlinearity, the required sampling frenquence for the output signal of amplifier is times of the original signal. The undersampling predistortion method can reduce the sampling frequency requirements. However, for the broadband signal, due to the significant memory effect of the power amplifier, the performance of undersampling digital predistortion is not satisfactory. In order to improve the undesirable predistortion performamce, this paper presents two kinds of undersampling digital predistortion compensation methods. For different sampling frequencies, the predistortion performances of these two kinds of compensation methods are different.