Nonnegative matrix factorization with Log Gabor wavelets for image representation and classification

Many problems in image representation and classification involve some form of dimensionality reduction. Nonnegative matrix factorization （NMF） is a recently proposed unsupervised procedure for learning spatially localized, partsbased subspace representation of objects. An improvement of the classical NMF by combining with Log-Gabor wavelets to enhance its part-based learning ability is presented. The new method with principal component analysis （PCA） and locally linear embedding （LIE） proposed recently in Science are compared. Finally, the new method to several real world datasets and achieve good performance in representation and classification is applied.

NECESSARY AND SUFFICIENT CONDITIONS FOR EXPANSIONS OF GABOR TYPE

In this paper, we give necessary and sufficient conditions for two families of Gabor functions of a certain type to yield a reproducing identity on L^2（R^n）. As applications, we characterize when such families yield orthonormal or bi-orthogonal expansions. We also obtain a generalization of the Balian-Low theorem for general reprodueing identities （not necessary coming from a frame）.

Properties of an improved Gabor wavelet transform and its applications to seismic signal processing and interpretation

This paper presents an analytical study of the complete transform of improved Gabor wavelets （IGWs）, and discusses its application to the processing and interpretation of seismic signals. The complete Gabor wavelet transform has the following properties. First, unlike the conventional transform, the improved Gabor wavelet transform （IGWT） maps time domain signals to the time-frequency domain instead of the time-scale domain. Second, the IGW＇s dominant frequency is fixed, so the transform can perform signal frequency division, where the dominant frequency components of the extracted sub-band signal carry essentially the same information as the corresponding components of the original signal, and the sub- band signal bandwidth can be regulated effectively by the transform＇s resolution factor. Third, a time-frequency filter consisting of an IGWT and its inverse transform can accurately locate target areas in the time-frequency field and perform filtering in a given time-frequency range. The complete IGW transform＇s properties are investigated using simulation experiments and test cases, showing positive results for seismic signal processing and interpretation, such as enhancing seismic signal resolution, permitting signal frequency division, and allowing small faults to be identified.

Global motion estimation with Gabor wavelet transform

A novel algorithm of global motion estimation is proposed. First, through Gabor wavelet transform （GWT）, a kind of energy distribution of image is obtained and checkpoints are selected according to a probability decision approach proposed. Then, the initialized motion vectors are obtained via a hierarcbal block-matching based on these checkpoints. Finally, by employing a 3-parameter motion model, precise parameters of global motion are found. From the experiment, the algorithm is reliable and robust.

Texture Image Classification Based on Gabor Wavelet

For a texture image, by recognizining the class of every pixel of the image, it can be partitioned into disjoint regions of uniform texture. This paper proposed a texture image classification algorithm based on Gabor wavelet. In this algorithm, characteristic of every image is obtained through every pixel and its neighborhood of this image. And this algorithm can achieve the information transform between different sizes of neighborhood.Experiments on standard Brodatz texture image dataset show that our proposed algorithm can achieve good classification rates.

Novel Facial Features Segmentation Algorithm

An efficient algorithm for facial features extractions is proposed. The facial features we segment are the two eyes, nose and mouth. The algorithm is based on an improved Gabor wavelets edge detector, morphological approach to detect the face region and facial features regions, and an improved T-shape face mask to locate the extract location of facial features. The experimental results show that the proposed method is robust against facial expression, illumination, and can be also effective if the person wearing glasses, and so on.

An Image Analysis of Breast Thermograms

Thermography （infrared imaging） is a non-invasive technique applied for the detection breast cancer. We consider the problem of automatically recognition malignant breast from frontal view thermography image presented as gray scale image. This framework provides insights into several issues： breast Region of Interest （ROI） detection, extraction statistical features, extraction features based on texture and co-occurrence matrix.

Characteristics of transmitting channel wave in a coal seam

An embedded underground coal seam carries channel waves of low seismic velocity along a stratigraphic rock-coal-rock sequence.In a homogeneous and isotropic seam, seismic waves propagate as trapped waves within the seam, which leads to propagation of channel waves.We describe how to set up a field test for transmission in order to acquire channel waves in a coal seam.Because channel wave signals are non-stationary in their frequencies and amplitudes, a necessary velocity spectrum and wavelet transformation analysis are applied to interpret the characteristics of channel waves.The advantage of using a wavelet transformation is that different resolutions can be obtained at different times and different frequencies.According to analysis of the seismic signals acquired in the S7 sensor hole, it was clearly shown that the characteristics of channel waves are lower frequencies and attenuation which can guide an effective wave for detecting voids, boundaries and faults in coal seams with strong roofs and floors.

Dissolution rate determination of alumina in molten cryolite-based aluminum electrolyte

Determination of dissolution rate of alumina is one of the classical problems in aluminum electrolysis. A novel method which can measure the dissolution rate of alumina was presented. Effect of factors on dissolution rate of alumina was studied intuitively and roundly using transparent quartz electrobath and image analysis techniques. Images about dissolution process of alumina were taken at an interval of fixed time from transparent quartz electrobath of double rooms. Gabor wavelet transforms were used for extracting and describing the texture features of each image. After subsampling several times, the dissolution rate of alumina was computed using these texture features in local neighborhood of samples. Regression equation of the dissolution rate of alumina was obtained using these dissolution rates. Experiments show that the regression equation of the dissolution rate of alumina is y=-0.000 5x^3＋0.024 0x^2-0.287 3x＋ 1.276 7 for Na3AIF6-AIF3-Al2O3-CaF2-LiF- MgF2 system at 920 ℃.

Robust visual tracking algorithm based on Monte Carlo approach with integrated attributes

To improve the reliability and accuracy of visual tracker,a robust visual tracking algorithm based on multi-cues fusion under Bayesian framework is proposed.The weighed color and texture cues of the object are applied to describe the moving object.An adjustable observation model is incorporated into particle filtering,which utilizes the properties of particle filter for coping with non-linear,non-Gaussian assumption and the ability to predict the position of the moving object in a cluttered environment and two complementary attributes are employed to estimate the matching similarity dynamically in term of the likelihood ratio factors;furthermore tunes the weight values according to the confidence map of the color and texture feature on-line adaptively to reconfigure the optimal observation likelihood model,which ensured attaining the maximum likelihood ratio in the tracking scenario even if in the situations where the object is occluded or illumination,pose and scale are time-variant.The experimental result shows that the algorithm can track a moving object accurately while the reliability of tracking in a challenging case is validated in the experimentation.

Recognition of human face based on improved multi-sample

In order to solve the problem caused by variation illumination in human face recognition,we bring forward a face recognition algorithm based on the improved multi-sample. In this algorithm,the face image is processed with Retinex theory,meanwhile,the Gabor filter is adopted to perform the feature extraction. The experimental results show that the application of Retinex theory improves the recognition accuracy,and makes the algorithm more robust to the variation illumination. The Gabor filter is more effective and accurate for extracting more useable facial local features. It is proved that the proposed algorithm has good recognition accuracy and it is stable under variation illumination.

Acoustic Emission Source Linear Localization Based on an Ultra-Short FBGs Sensing System

An acoustic emission （AE） linear location system was proposed, which employed fiber Bragg gratings （FBGs） as AE sensors. It was demonstrated that the FBG wavelength could be modulated as the static case when the grating length was much shorter than the AE wavelength. In addition, an improved AE location method based on the Gabor wavelet transform （WT） and threshold analysis was represented. The method was testified through AE linear location experiments based on a tunable narrow-band laser interrogation system using ultra-short FBG sensors as AE sensors. Results of the experiments showed that 86% of the linear location errors were less than 10mm.