Wavelet Moments Research Articles

An Intelligent Fused Approach for Face Recognition

AbstractFace detection plays important roles in many applications such as human–computer interaction, security and surveillance, face recognition, etc. This article presents an intelligent enhanced fused approach for face recognition based on the Voronoi diagram (VD) and wavelet moment invariants. Discrete wavelet transform and moment invariants are used for feature extraction of the facial face. Finally, VD and the dual tessellation (Delaunay triangulation, DT) are used to locate and detect original face images. Face recognition results based on this new fusion are promising in the state of the art.

3D B Spline Interval Wavelet Moments for 3D Objects

By introducing B spline interval wavelets to the 3D wavelet moments, we give a new representation for 3D objects. The new descriptors can keep more information of the edge. They are invariant to translation, rotation and scale, and have the multi-resolution features in the radial direction, which can handle noise to some extent and provide multi-level features to satisfy various requirements. With the Pearson correlation coefficient and the energy in each level, we give a new method to measure the similarity of two objects. In the experiment, it is proved to be an efficient method.

Method of Feature Extraction Based on Wavelet and Affine Invariant Moment for Target Recognition

Method of Feature Extraction Based on Wavelet and Affine Invariant Moment for Target Recognition

Face recognition using complex wavelet moments

The wavelet moments are very useful image descriptors which contain both global and local characteristics of an image. Its usefulness is further characterized by rotation invariant property of its magnitude. The phase information is, however, left out because it changes with image rotation. In this paper, we incorporate phase information and use both the real and imaginary components of wavelet moments to describe an image and develop a similarity measure which is invariant under image rotation. The proposed framework is applied to face recognition. Extensive experimental results are provided to demonstrate the enhanced performance of the proposed framework as compared to the magnitude only framework. The results are also compared with many other state-of-the-art techniques used for face recognition.

Wavelet Spectrum for Investigating Statistical Characteristics of UDP-Based Internet Traffic

In this paper, we consider statistical characteristics of real User Datagram Protocol (UDP) traffic. Four main issues in the study include(i) the presence of long rangedependence (LRD) in the UDP traffic,(ii) the marginal distribution of the UDP traces,(iii) dependence structure of wavelet coefficients,(iv) and performance evaluation of the Hurst parameter estimation based on different numbers of vanishing moments of the mother wavelet . By analyzing a large set of real traffic data, it is evident that the UDP Internet traffic reveals the LRD properties with considerably high non-stationary processes.Furthermore, it exhibits non-Gaussian marginal distributions. However, by increasing the number of vanishing moments,it is impossible to achieve reduction fromLRD to become a short range dependence. Thus, it can be shown that there is no significant difference in performance estimation of the Hurst parameter for different numbers of vanishing moments of the mother wavelet.

Performance analysis of various local and global shape descriptors for image retrieval

In this paper, various prominent local and global descriptors are evaluated against each other for analyzing their performance on shape-based image retrieval. Local descriptors include Fourier descriptors, Weber's local descriptor, local binary patterns, and local ternary patterns. The prominent global descriptors include moment invariants, generic Fourier descriptor (GFD), angular radial transform (ART), wavelet moments (WM), and Zernike moment descriptor (ZMD). In addition, a novel local descriptor is proposed based on the histograms of circular arcs and linear edges, which are detected by means of Hough transform. The proposed local descriptor provides features, which are invariant to geometric transformations and are robust to noise as compared to some existing prominent local descriptors. We also propose an improvement in the performance of global descriptors GFD, ART, WM, and ZMD by taking advantage of the phase information in the comparison process along with their magnitude. Subsequently, the local and global descriptors with the best image-retrieval performances are combined to design an effective retrieval system, which further enhances the retrieval performance substantially. All descriptors are analyzed in terms of six principles set by MPEG-7. Detailed experiments are performed on standard benchmark image databases along with their rotation-invariance and noise test. The results of experiments reveal that the proposed fusion of local and global descriptors outperforms other major descriptors.

Discrimination ability improvement of invariant moment based on multi-scale mathematical morphology

Discrimination ability of invariant moment is crucial for similar object recognition and classification. To improve the discrimination ability of invariant moment for similar object classification, a multi-scale mathematical morphology based algorithm is proposed. Firstly, the multi-scale operations are used to extract the details of image at different scales. Then, the details of multi-scales are combined to form the very different part of the similar objects, and the corresponding invariant moments of the very different parts are calculated. Finally, a measure which is similar to the between-to within-class variance ratio is used to select the efficient invariant moments of the very different parts of several scales. The selected moments of the very different parts of the similar objects will be more efficient than the original invariant moments of original images. Hu moment is adopted as an example to show the efficiency of improvement. And, any type of invariant moment can be used to replace Hu moment in our algorithm to achieve a more efficient invariant moment algorithm. So, our proposed algorithm is very important and can be widely used in different similar object discrimination. The application of improved Hu moment by our algorithm are compared with the original Hu moment and wavelet moment, experimental results show that our algorithm is very efficient.

Simultaneous Quantitative Analysis of Three Compounds Using Three-Dimensional Fluorescence Spectra Based on Digital Image Techniques

Digital image processing has been applied on various fields such as classification and qualitative analysis. In this work, a very simple quantitative approach was proposed for the first time. Based on the digital grayscale images of three-dimensional fluorescence spectra, several wavelet moment invariants were calculated, and used to establish the linear models for the quantitative analysis. This approach was applied to the quantitative analysis of Tryptophan, Tyrosine and Phenylalanine in mixture samples, and the correlation coefficients R(2) of the obtained linear models were more than 0.99, which were supported by the strict statistical parameters as well as leave-one-out and Jackknife cross-validations. Our study indicates that the selected wavelet moment invariants are immune from the noise and background signals, and the quantitative analysis can be performed accurately based on the overlapping peaks of compounds in mixture. This proposed approach provides a novel pathway for the analysis of three-dimensional spectra.

Efficient Face Recognition System using Artificial Neural Network

facial feature is needed to build a robust face recognition system capable of suppress the effect of illumination and pose variation. In this paper, a robust face recognition system is proposed. In the proposed system, two level haar wavelet transform is used to decompose frontal face image into seven sub-image bands. Thereafter eigenface feature is extracted from these bands. The feature is used as input to the classification algorithm based on Back Propagation Neural Network (BPNN). The proposed system has been tested using 150 frontal face samples with illumination and pose variation. The results obtained are very encouraging. level image resolution before extract moments features from face image while in (10) they presented an enhanced fused approach for face recognition based on voronoi diagram and wavelet moment invariants. Discrete wavelet transform and moment invariants are used for feature extraction of the facial face. Gabor filter, log Gabor filter and discrete wavelet transform are applied to face image prior to dimensionality reduction by Principal Component Analysis (PCA) in the paper presented in (11). The approximation coefficients in discrete wavelet transform is extracted and it was used to compute the face recognition accuracy instead of using all the coefficients. In (12) they proposed Face Recognition Algorithms that uses combination of Gabor and eigenfaces to obtain the feature vector. This feature vector is feed to a multilayer perceptron to carry out the face recognition or identity verification tasks. The proposed system is meant to provide robustness against changes in illumination and facial expressions. The paper presented in (13) is used tests the performance of Principle Component Analysis (PCA) based technique for face recognition. Recognition of human faces with two facial expressions is done. Classification of face images are carried out using distance based classifier and Neural Network (NN). In this work, better method for face recognition is proposed using our database. Eigenface feature is extracted from sub- band images of the second level Haar wavelet transform decomposition product. Neural Network is used as the classification algorithm. The feature is robust enough to suppress the effect of varying pose and illumination. The rest of the paper is organized as follows: Section 2 describes the system component. Face image decomposition is presented in section 3, section 4 introduces the feature extraction method, and section 5 describes classification process based on Neural Network. The experimental results are shown in section 6. Finally, conclusion is presented in section 7.

An Obstacle Recognition Method Based on Improved PSO-WNN for Deicing Robot on Voltage Transmission Line

An intelligent obstacle recognition method based on improved particle swarm optimization wavelet neural network (IPSO-WNN) for deicing robot on voltage transmission line is presented because traditional obstacle recognition methods often have errors due to structural constraints. Firstly, the improved PSO algorithm is proposed by introduce idea of probabilistic leap in simulated annealing. Secondly , the improved algorithm is used to optimize the parameters of the wavelet network. Finally, the trained IPSO-WNN is used to recognize and classify the wavelet moments of the obstacle edged images of robot. Experimental results show that compared with PSO-WNN this method have quicker convergence rate and higher classification precision and improve obviously recognition function of obstacle on transmission line.

Identification of complex sound sources produced by gear units

Noise source visualisation represents an important tool for technical acoustics. Many techniques of noise source visualisation have been developed, based on a specific noise source in a specific type of acoustic environment. A new visualisation method of complex noise sources is presented, using an acoustic camera and a new algorithm. Different transient acoustical phenomena can be noted. Additionally, a new family of biorthogonal wavelets is applied to determine fault in gears. The new wavelets are a generalisation of biorthogonal wavelet systems. Smoothness is controlled independently in the analysis. For the optimisation of the synthesis bank, discrete finite variation is used. Differentiability is measured, for which a large number of vanishing wavelet moments is necessary, in favour of a smoothness measure based on the fact that a finite depth of the filter bank tree is in most case related to practical applications.

On the Hilbert Transform of Wavelets

A wavelet is a localized function having a prescribed number of vanishing moments. In this correspondence, we provide precise arguments as to why the Hilbert transform of a wavelet is again a wavelet. In particular, we provide sharp estimates of the localization, vanishing moments, and smoothness of the transformed wavelet. We work in the general setting of non-compactly supported wavelets. Our main result is that, in the presence of some minimal smoothness and decay, the Hilbert transform of a wavelet is again as smooth and oscillating as the original wavelet, whereas its localization is controlled by the number of vanishing moments of the original wavelet. We motivate our results using concrete examples.

A robust digital audio watermarking scheme using wavelet moment invariance

It is a challenging work to design a robust audio watermarking scheme against various attacks. Wavelet moment invariances are new features combining the moment invariant features and the wavelet features, and they have some excellent characteristics, such as the ability to capture local information, robustness against common signal processing, and the linear relationship between a signal and its wavelet moments etc. Based on wavelet moment and synchronization code, we propose a new digital audio watermarking algorithm with good auditory quality and reasonable resistance against most attacks in this paper. Firstly, the origin digital audio is segmented and then each segment is cut into two parts. Secondly, with the spatial watermarking technique, synchronization code is embedded into the statistics average value of audio samples in the first part. And then, map 1D digital audio signal in the second part into 2D form, and calculate its wavelet moments. Finally, the watermark bit is embedded into the average value of modulus of the low-order wavelet moments. Meanwhile combining the two adjacent synchronization code searching technology, the algorithm can extract the watermark without the help from the origin digital audio signal. Simulation results show that the proposed watermarking scheme is not only inaudible and robust against common signals processing such as MP3 compression, noise addition, resampling, and re-quantization etc., but also robust against the desynchronization attacks such as random cropping, amplitude variation, pitch shifting, and jittering etc.

A Gait Recognition Algorithm Based on Wavelet Moment and Double Triangle Feature

In order to improve the classification rate of gait recognition, a new gait recognition algorithm is proposed. Firstly, the gait images are preprocessed, and the outlines of gait images are extracted and normalized. Secondly, wavelet moments of the outlines are calculated to describe the static feature of the gait images. Thirdly, the leg double triangle model is built. The first triangle consists of the mid-point of the two hips, left knee point and right knee point, and the other one consists of the mid-point of the two hips, left ankle point and right ankle point. Then the parameters of two triangles are extracted to describe the dynamic features of the gait images. Finally, the above two features are fused and used for the classification. The experimental results show that proposed algorithm provides higher correct classification rate than the algorithms using single feature, and meets the requirements of the real-time.

Automatic Modulation Recognition Using Wavelet Transform and Neural Networks in Wireless Systems

Modulation type is one of the most important characteristics used in signal waveform identification. In this paper, an algorithm for automatic digital modulation recognition is proposed. The proposed algorithm is verified using higher-order statistical moments (HOM) of continuous wavelet transform (CWT) as a features set. A multilayer feed-forward neural network trained with resilient backpropagation learning algorithm is proposed as a classifier. The purpose is to discriminate among different M-ary shift keying modulation schemes and themodulation order without any priori signal information. Pre-processing and features subset selection using principal component analysis is used to reduce the network complexity and to improve the classifier's performance. The proposed algorithm is evaluated through confusion matrix and false recognition probability. The proposed classifier is shown to be capable of recognizing the modulation scheme with high accuracy over wide signal-to-noise ratio (SNR) range over both additive white Gaussian noise (AWGN) and different fading channels.

Construction of optimally conditioned cubic spline wavelets on the interval

The paper is concerned with a construction of new spline-wavelet bases on the interval. The resulting bases generate multiresolution analyses on the unit interval with the desired number of vanishing wavelet moments for primal and dual wavelets. Both primal and dual wavelets have compact support. Inner wavelets are translated and dilated versions of well-known wavelets designed by Cohen, Daubechies, and Feauveau. Our objective is to construct interval spline-wavelet bases with the condition number which is close to the condition number of the spline wavelet bases on the real line, especially in the case of the cubic spline wavelets. We show that the constructed set of functions is indeed a Riesz basis for the space L 2 ([0, 1]) and for the Sobolev space H s ([0, 1]) for a certain range of s. Then we adapt the primal bases to the homogeneous Dirichlet boundary conditions of the first order and the dual bases to the complementary boundary conditions. Quantitative properties of the constructed bases are presented. Finally, we compare the efficiency of an adaptive wavelet scheme for several spline-wavelet bases and we show a superiority of our construction. Numerical examples are presented for the one-dimensional and two-dimensional Poisson equations where the solution has steep gradients.

Palmprint matching algorithm based on-àtrous-Contourlet transform and invariant moments

In order to improve the speed and accuracy of palmprint identification system and overcome the weakness of Contourlet transform in dealing with high-dimensional signals,a novel identification algorithm of palmprint feature was proposed.Firstly,the atrous-Contourlet transformation for palmprint image was decomposed to obtain high frequency coefficients and low frequency coefficients in the different sub-band of different direction.Then,different feature weighting coefficients according to the statistical characteristics extracted from different frequency domains were chosen,which were calculated to get some new invariant moments vectors.Finally,the palmprint identification was achieved.The experimental results show that the method here is more effective in matching than wavelet moment algorithm,Hu's invariant moment algorithm and Contourlet algorithm.

Outer Space Target Recognition Based on Image Sequence

Outer space target recognition is one of kernel functions of space-based surveillance system. This paper presents a new method on space target recognition based on image sequence using wavelet moments and DS evidence theory. Experimental results show our recognition system has a good performance (100%) which achieves need of practical application.

Urban land cover classification from very high resolution imagery using spectral and invariant moment shape information

Detailed urban land cover classification using very high resolution (VHR) imagery is a challenging task, partly due to the spectral similarity between land cover classes and high spectral variability within individual classes. In this paper, shape features extracted from segmented objects using three types of invariant moments, namely Hu’s moments, Zernike moments, and wavelet moments, were applied to urban land cover classification using VHR imagery. The performance of invariant moments was evaluated by comparison with the performance of the object-based classification utilizing spectral information alone. The results showed that overall accuracy improved significantly when shape features defined by invariant moments were combined with spectral bands in urban land cover classification. In particular, significant improvements in accuracy for individual classes were achieved for those classes which have similar spectral features but different shape features, such as building, road, and other impervious surface. The study also found that selection of segmentation results is crucial in extraction of shape features for improved classification.

Adaptive Nonseparable Wavelet Transform via Lifting and its Application to Content-Based Image Retrieval

We present in this paper a novel way to adapt a multidimensional wavelet filter bank, based on the nonseparable lifting scheme framework, to any specific problem. It allows the design of filter banks with a desired number of degrees of freedom, while controlling the number of vanishing moments of the primal wavelet ((~)N moments) and of the dual wavelet ( N moments). The prediction and update filters, in the lifting scheme based filter banks, are defined as Neville filters of order (~)N and N, respectively. However, in order to introduce some degrees of freedom in the design, these filters are not defined as the simplest Neville filters. The proposed method is convenient: the same algorithm is used whatever the dimensionality of the signal, and whatever the lattice used. The method is applied to content-based image retrieval (CBIR): an image signature is derived from this new adaptive nonseparable wavelet transform. The method is evaluated on four image databases and compared to a similar CBIR system, based on an adaptive separable wavelet transform. The mean precision at five of the nonseparable wavelet based system is notably higher on three out of the four databases, and comparable on the other one. The proposed method also compares favorably with the dual-tree complex wavelet transform, an overcomplete nonseparable wavelet transform.