Advanced Materials Research Vols. 998-999

Abstract: Pattern matching algorithm is widely used. It plays an important role in information retrieval, data mining, intrusion detection and other fields. Among them, the BM algorithm is the most common. A new improved algorithm-BMQ algorithm is proposed on the basis of BM and related algorithms. The improved algorithm makes use of uniqueness and combination of the last character and next character of string, to increase the probability of the maximum right shift. Theoretical analysis and experimental comparison shows that the BMQ is better than BM algorithms in the process of string matching and string searching; in order to further verify its effectiveness, the improved algorithm is introduced to intrusion detection system, the experimental results show that BMQ algorithm improves the efficiency of intrusion detection.

Abstract: Images taken by wide-angle cameras tend to have severe distortions which pull points towards the optical center. Matlab calibration toolbox is widely used in computer vision for its simplicity and strong interactivity; however it cannot accurately extract corners from wide-angle image, and be unable to correct image automatically. This paper proposes an improved wide-angle image correction method which can obtain corners precisely and implement correction automatically. The experimental results demonstrate this method not only improves the accuracy of corner extracting, but also makes the wide-angle image correction automated.

Abstract: This paper presented a detecting algorithm based on improved Gaussian mixture model, which improved the speed of establishing and updating the model. Whilst, according to the detection requirements in different time periods, using the method of combining temporal differencing and background differencing to detect moving body improved the instantaneity and veracity of moving body detection.

Abstract: The wavelet threshold denoising method is widely researched and used, the hard threshold and soft threshold which exist different advantages and disadvantages are two kinds commonly used threshold function; the threshold function was improved between hard threshold and soft threshold by using the exponential function. The smooth curve is monotonically increasing, no discontinuity and thus will not produce oscillation. Overcoming shortcomings that the hard threshold function and soft threshold function to be zero in wavelet coefficient is small and soft threshold exist constant deviation, and when the absolute value is greater than the threshold coefficient, soft threshold function is disposed with the same contraction method. After the simulation results prove that compared with traditional speech enhancement algorithms, the new threshold function has been improved to a certain extent.

Abstract: Based on Hilbert-Huang Transform (HHT) theory, we present a method to analyze the electroencephalogram (EEG) signal of right and left hand motor imagery. Firstly, EMD method decomposed EEG signal into a group of intrinsic mode functions (IMFs). The first three IMFs were extracted to denoise. We adopt endpoint Mirror Extension method to relieve the influence on subsequent processing brought by endpoint effect. According to the Hilbert transform, we can obtain the time-frequency distribution. The energy of the first three components is selected as the input of SVM. The results show that EMD is an efficient method to analyze the EEG signal. The proposed method obtains an ideal recognition rate.

Abstract: Based on the shortcomings of standard median filtering and combined with the mean filtering, this paper puts forward two improved median filtering algorithms referred as the weighted fast median filtering algorithm and the weighted adaptive median filtering algorithm. The experiment results with MATLAB show that weighted fast median filtering algorithm has a significant effect on low-density impulse noise, and accelerates the speed of real-time processing. The weighted median filter could effectively eliminate the high-density impulse noise from polluted images, and better maintains the details of the original image.

Abstract: Association rules extraction is one of the important goals of data mining and analyzing. Aiming at the problem that information lose caused by crisp partition of numerical attribute , in this article, we put forward a fuzzy association rules mining method based on fuzzy logic. First, we use c-means clustering to generate fuzzy partitions and eliminate redundant data, and then map the original data set into fuzzy interval, in the end, we extract the fuzzy association rules on the fuzzy data set as providing the basis for proper decision-making. Results show that this method can effectively improve the efficiency of data mining and the semantic visualization and credibility of association rules.

Abstract: In the face of the carrier signal is discrete spectrum analysis, frequency, amplitude and phase are likely to produce the error problem. Based on discrete spectrum correction theory, derivation of the carrier signal causes the error when the discrete spectrum analysis and using the generalized cosine window for the error of the discrete spectrum analysis.

Abstract: Image processing has been widely applied in various fields. Photoshop image processing software is the best one for the enthusiasts. At present, more and more people are interested in using Photoshop to create some characteristic pictures, while the key and difficulty part of image processing in Photoshop is matting method (create the image area).In order to help the image processing enthusiasts to accomplish in the process of image processing and create a more beautiful image, this paper introduces several commonly used matting methods in Photoshop and discusses in detail the scope of application, the using methods, the advantages and disadvantages of them.

Abstract: A method for driver fatigue detection based on eye locating was researched in this paper.. The eye location was achieved by combining gray information with shape information, and matched the eye template of image with which was in the open state. To observe images within a certain time interval was to identify the open or closed state of the drivers' eyes, so as to determine if they have fatigue driving. The results showed that the algorithm could suppress gaussian noise and impulse noise very effectively, and had better filtering performance than the standard median filters..