MOVING TARGETS PATTERN RECOGNITION BASED ON THE WAVELET NEURAL NETWORK

Based on pattern recognition theory and neural network technology, moving objects automatic detection and classification method integrating advanced wavelet analysis are discussed in detail. An algorithm of moving targets pattern recognition on the combination of inter-frame difference and wavelet neural network is presented. The experimental results indicate that the designed BP wavelet network using this algorithm can recognize and classify moving targets rapidly and effectively.

Application of Wavelet Analysis toInterference Elimination for Geochemical Hydrocarbon Exploration

Interference in the data of geochemical hydrocarbon exploration is a large obstacle for anomaly recognition. The multiresolution analysis of wavelet analysis can extract the information at different scales so as to provide a powerful tool for information analysis and processing. Based on the analysis of the geometric nature of hydrocarbon anomalies and background, Mallat wavelet and symmetric border treatment are selected and data pre-processing (logarithm-normalization) is established. This approach provide good results in Shandong and Inner Mongolia, China. It is demonstrated that this approach overcome the disadvantage of backgound variation in the window (interference in window), used in moving average, frame filtering and spatial and scaling modeling methods.

New pattern recognition system in the e-nose for Chinese spirit identification

This paper presents a new pattern recognition system for Chinese spirit identification by using the polymer quartz piezoelectric crystal sensor based e-nose. The sensors are designed based on quartz crystal microbalance（QCM） principle,and they could capture different vibration frequency signal values for Chinese spirit identification. For each sensor in an8-channel sensor array, seven characteristic values of the original vibration frequency signal values, i.e., average value（A）,root-mean-square value（RMS）, shape factor value（S_f）, crest factor value（C_f）, impulse factor value（I_f）, clearance factor value（CL_f）, kurtosis factor value（K_v） are first extracted. Then the dimension of the characteristic values is reduced by the principle components analysis（PCA） method. Finally the back propagation（BP） neutral network algorithm is used to recognize Chinese spirits. The experimental results show that the recognition rate of six kinds of Chinese spirits is 93.33% and our proposed new pattern recognition system can identify Chinese spirits effectively.

Evaluation of echo features of ultrasonic flaws and its intelligent pattern recognition

In this paper, three types of weld flaw were taken as target, evaluation and recognition of flaw echo features were studied. On the basis of experimental study and theoretical analysis, 26 features have been extracted from each echo samples. A method which is based on the xtatislical hypothesis testing and used for feature evaluation and optimum subset selection was explored. Thus, the dimensionality reduction of feature space was brought out, and simultaneously the amount of calculation was decreased. An intelligent pattern classifier with B-P type neural network was constructed which was characterized by high speed and accuracy for learning. Using a half of total samples as training set and others as testing set, the learning efficiency and the classification ability of network model were studied. The results of experiment showed that the learning rate of different training samples was about 100%. The results of recognition was satisfactory when the optimum feature subset was taken as the sample's feature vectors. The average recognition rate of three type flaws was about 87.6%, and the best recognition rate amounted to 97%.

A systematic method based on statistical pattern recognition for estimating product quality on-line

To avoid the complexity of building mechanistic models by studying the inner nature of the object, a systematic method based on statistical pattern recognition is developed in order to estimate the product quality on-line. The mapping relationship between a feature space and a product quality space can be built by using regression analysis, and in applying clustering analysis the product quality space can be partitioned automatically. Eventually, estimating product quality on-line can be accomplished by sorting the mapped data in the partitioned quality space. A concrete problem is proposed which has a relatively small ratio of training data to input variables. By implementing the method mentioned above, a satisfying result has been achieved. Furthermore, the further question about choosing suitable mapping methods is briefly discussed.

Medical Image Segmentation Based on Wavelet Analysis and Gradient Vector Flow

Medical image segmentation is one of the key technologies in computer aided diagnosis. Due to the complexity and diversity of medical images, the wavelet multi-scale analysis is introduced into GVF (gradient vector flow) snake model. The modulus values of each scale and phase angle values are calculated using wavelet transform, and the local maximum points of modulus values, which are the contours of the object edges, are obtained along phase angle direction at each scale. Then, location of the edges of the object and segmentation is implemented by GVF snake model. The experiments on some medical images show that the improved algorithm has small amount of computation, fast convergence and good robustness to noise.

Application of wavelet transform in feature extraction and pattern recognition of wideband echoes

A novel approach to extract edge features from wideband echo is proposed. The set of extracted features not only represents the echo waveform in a concise way, but also is sufficient and well suited for classification of non-stationary echo data from objects with different property.The feature extraction is derived from the Discrete Dyadic Wavlet Transform (DDWT) of the echo through the undecimated algorithm. The motivation we use the DDWT is that it is time-shift-invariant which is beneficial for localization of edge, and the wavelet coefficients at larger scale represent the main shape feature of echo, i.e. edge, and the noise and modulated high-frequency components are reduced with scale increased. Some experimental results using real data which contain 144 samples from 4 classes of lake bottoms with different sediments are provided. The results show that our approach is a prospective way to represent wideband echo for reliable recognition of nonstationary echo with great variability.

Regeneration pattern analysis of Quercus liaotungensis in a temperate forest using two-dimensional wavelet analysis

This paper introduces the two-dimensional(2D)wavelet analysis as a general interrogative technique for the detection of spatial structure in lattice data.The 2D wavelet analysis detects components of hierarchical structure and displays the locational information of the components.Patches and gaps of different spatial scales in graphical presentation of wavelet coefficients can be linked to the local ecological processes that determine patterns at stand or landscape scales.Derived from the 2D wavelet transform function,the calculation of wavelet variance can reduce the four-dimensional data of wavelet coefficients to a two-dimensional wavelet variance function and quantify the contribution of the given scale to the overall pattern.We illustrate the use of the 2D wavelet analysis by analyzing two simulated patterns and identifying the regeneration pattern of the Quercus liaotungensis in a warm temperate forest in north China.Our results indicate that the recruitment of Q.liaotungensis occurs in an overlapping area between the patch of adult and canopy gap at scales of 45m×45m–70m×70m and 20m×20m–30m×30m.The regeneration pattern of Q.liaotungensis can be mainly ascribed to a trade-off between two ecological processes:recruitment around parent trees and the physiological light requirements of seedlings and saplings.Our results provide a general portrayal of the regeneration pattern for the dispersal-limited and shade-intolerant Quercus species.We find that the two-dimensional wavelet analysis efficiently characterizes the scale-specific pattern of Q.liaotungensis at different life-history stages.

An unsupervised pattern recognition methodology based on factor analysis and a genetic-DBSCAN algorithm to infer operational conditions from strain measurements in structural applications

Structural Health Monitoring(SHM) suggests the use of machine learning algorithms with the aim of understanding specific behaviors in a structural system. This work introduces a pattern recognition methodology for operational condition clustering in a structure sample using the well known Density Based Spatial Clustering of Applications with Noise(DBSCAN) algorithm.The methodology was validated using a data set from an experiment with 32 Fiber Bragg Gratings bonded to an aluminum beam placed in cantilever and submitted to cyclic bending loads under 13 different operational conditions(pitch angles). Further, the computational cost and precision of the machine learning pipeline called FA + GA-DBSCAN(which employs a combination of machine learning techniques including factor analysis for dimensionality reduction and a genetic algorithm for the automatic selection of initial parameters of DBSCAN) was measured. The obtained results have shown a good performance, detecting 12 of 13 operational conditions, with an overall precision over 90%.

Finger crease pattern recognition using Legendre moments and principal component analysis

The finger joint lines defined as finger creases and its distribution can identify a person. In this paper, we propose a new finger crease pattern recognition method based on Legendre moments and principal component analysis （PCA）. After obtaining the region of interest （ROI） for each finger image in the pre- processing stage, Legendre moments under Radon transform are applied to construct a moment feature matrix from the ROI, which greatly decreases the dimensionality of ROI and can represent principal components of the finger creases quite well. Then, an approach to finger crease pattern recognition is designed based on Karhunen-Loeve （K-L） transform. The method applies PCA to a moment feature matrix rather than the original image matrix to achieve the feature vector. The proposed method has been tested on a database of 824 images from 103 individuals using the nearest neighbor classifier. The accuracy up to 98.584% has been obtained when using 4 samples per class for training. The experimental results demonstrate that our proposed approach is feasible and effective in biometrics.

Pattern recognition of surface electromyography signal based on wavelet coefficient entropy

This paper introduced a novel, simple and ef-fective method to extract the general feature of two surface EMG (electromyography) signal patterns: forearm supination (FS) surface EMG signal and forearm pronation (FP) surface EMG signal. After surface EMG (SEMG) signal was decomposed to the fourth resolution level with wavelet packet transform (WPT), its whole scaling space (with frequencies in the interval (0Hz, 500Hz]) was divided into16 frequency bands (FB). Then wavelet coefficient entropy (WCE) of every FB was calculated and corre-spondingly marked with WCE(n) (from the nth FB, n=1,2,…16). Lastly, some WCE(n) were chosen to form WCE feature vector, which was used to distinguish FS surface EMG signals from FP surface EMG signals. The result showed that the WCE feather vector consisted of WCE(7) (187.25Hz, 218.75Hz) and WCE(8) (218.75Hz, 250Hz) can more effectively recog-nize FS and FP patterns than other WCE feature vector or the WPT feature vector which was gained by the combination of WPT and principal components analysis.

Performance evaluation of high frequency sub-bands of wavelet transform for palmprint recognition

Wavelet decomposition has been applied in palmprint recognition successfully. However, only the low frequency sub-band was used for further feature extraction, while the high frequency sub-bands were considered to be unsuitable for palmprint recognition due to their sensitivity to noise and shape distortion. In this paper, we firstly investigate the performances of all the sub-bands by using principal component analysis (PCA) on the BJTU and PolyU palmprint databases, and then use mean filtering to enhance the robustness of the high frequency sub-bands. We find that the preprocessed high frequency sub-bands not only can be used for palmprint recognition but also contain complementary information with the low frequency sub-band. The experimental results show that the performances of the horizontal and vertical high frequency sub-bands can be promoted up to a competitive level, and the fusion scheme, which combines the matching scores of high frequency sub-bands with that of low frequency sub-band, is superior to the conventional recognition methods.

A Pattern Recognition Framework for Detecting Changes in Chinese Internet Management System

Past studies on the Chinese internet management system have revealed a smart internet management system that takes advantage of time to filter content with collective action potential.How and why such a system was institutionalized?We offer a historical institutional analysis to explain the way in which the system evolved.We implement social network analysis to examine the Weibo posts of recurrent events,the elections in Area A in 2016 and 2018,to identify pattern changes in the system.There are two aspects of the changes:the centralization of the command line to a single authority and the implementation of a discriminatory strategy to deal with the various online expressions together forming this intelligent system.The improved Chinese information surveillance system demonstrates both a top-down information management and a bottom-up opinion formation.

Modelling Animal Activity as Curves: An Approach Using Wavelet-Based Functional Data Analysis

Temporal activity patterns in animals emerge from complex interactions between choices made by organisms as responses to biotic interactions and challenges posed by external factors. Temporal activity pattern is an inherently continuous process, even being recorded as a time series. The discreteness of the data set is clearly due to data-acquisition limitations rather than a true underlying discrete nature of the phenomenon itself. Therefore, curves are a natural representation for high-frequency data. Here, we fully model temporal activity data as curves integrating wavelets and functional data analysis, allowing for testing hypotheses based on curves rather than on scalar and vector-valued data. Temporal activity data were obtained experimentally for males and females of a small-bodied marsupial and modelled as wavelets with independent and identically distributed errors and dependent errors. The null hypothesis of no difference in temporal activity pattern between male and female curves was tested with functional analysis of variance (FANOVA). The null hypothesis was rejected by FANOVA and we discussed the differences in temporal activity pattern curves between males and females in terms of ecological and life-history attributes of the reference species. We also performed numerical analysis that shed light on the regularity properties of the wavelet bases used and the thresholding parameters.

Novel Face Recognition Method by Combining Spatial Domain and Selected Complex Wavelet Features

A novel face recognition method based on fusion of spatial and frequency features was presented to improve recognition accuracy. Dual-Tree Complex Wavelet Transform derives desirable facial features to cope with the variation due to the illumination and facial expression changes. By adopting spectral regression and complex fusion technologies respectively, two improved neighborhood preserving discriminant analysis feature extraction methods were proposed to capture the face manifold structures and locality discriminatory information. Extensive experiments have been made to compare the recognition performance of the proposed method with some popular dimensionality reduction methods on ORL and Yale face databases. The results verify the effectiveness of the proposed method.

Automatic recognition and quantitative analysis of Ω phases in Al-Cu-Mg-Ag alloy

The main methods of the second phase quantitative analysis in current material science researches are manual recognition and extracting by using software such as Image Tool and Nano Measurer.The weaknesses such as high labor intensity and low accuracy statistic results exist in these methods.In order to overcome the shortcomings of the current methods,the Ω phase in Al-Cu-Mg-Ag alloy is taken as the research object and an algorithm based on the digital image processing and pattern recognition is proposed and implemented to do the Al alloy TEM(transmission electron microscope)digital images process and recognize and extract the information of the second phase in the result image automatically.The top-hat transformation of the mathematical morphology,as well as several imaging processing technologies has been used in the proposed algorithm.Thereinto,top-hat transformation is used for elimination of asymmetric illumination and doing Multi-layer filtering to segment Ω phase in the TEM image.The testing results are satisfied,which indicate that the Ω phase with unclear boundary or small size can be recognized by using this method.The omission of these two kinds of Ω phase can be avoided or significantly reduced.More Ω phases would be recognized(growing rate minimum to 2% and maximum to 400% in samples),accuracy of recognition and statistics results would be greatly improved by using this method.And the manual error can be eliminated.The procedure recognizing and making quantitative analysis of information in this method is automatically completed by the software.It can process one image,including recognition and quantitative analysis in 30 min,but the manual method such as using Image Tool or Nano Measurer need 2 h or more.The labor intensity is effectively reduced and the working efficiency is greatly improved.

基于HPLC指纹图谱结合化学模式识别对葡萄醋的质量评价

目的建立葡萄醋中有机酸成分的高效液相(High performance liquid phase,HPLC)指纹图谱,并结合化学模式识别对葡萄醋质量进行评价及为临床应用提供参考。方法采用HPLC法,以甲醇(A)∶0.5%磷酸水溶液(D)=1∶99洗脱,建立葡萄醋指纹图谱,同时测定琥珀酸的含量;采用相似度评价、聚类分析(Cluster analysis,CA)、主成分分析(Principal component analysis,PCA)和偏最小二乘-判别分析(Partial least squares discriminant analysis,PLS-DA)等化学模式识别对差异性特征成分进行筛选。结果15批葡萄醋样品的指纹图谱共有峰有11个,通过与混合对照品色谱峰进行比对,指认出6个共有峰,分别是酒石酸(1号峰),苹果酸(2号峰),琥珀酸(3号峰),富马酸(4号峰),绿原酸(9号峰),乙酸(11号峰);CA结果表明15批葡萄醋样品中各个化合物之间含量差异较大,即使是同一生产企业生产的不同批次葡萄醋其含量差异也较大;PCA结果显示葡萄醋共有峰含量差异的主要成分并非单一成分,而是前4种主成分共同的影响;PLS-DA结果表明筛选出的7、3、9、4号峰可作为鉴别和区分葡萄醋质量差异的标志物,其中包括共有峰3号峰(琥珀酸)、4号峰(富马酸)。结论采用指纹图谱结合化学模式识别技术可快速、有效地筛选出葡萄醋的差异性特征成分,为葡萄醋质量评价及临床应用提供参考。