Local Statistical Features Research Articles

Deconvolution image ringing artifacts removal via anisotropic diffusion

This paper describes a new anisotropic diffusion method for removing ringing artifacts from deconvolution images. The characteristics of the deconvolution image ringing artifact are also analyzed. Unlike the previous anisotropic diffusion methods that use the global statistical value as the scale function, the proposed anisotropic diffusion defines a new scale function based on image local statistical characteristics to recognize the ringing artifacts. The proposed anisotropic diffusion performs well in terms of the images' detail preservation, error reduction, and improvement in visual quality in the presence of various levels of image ringing artifacts.

Image Super-Resolution With Sparse Neighbor Embedding

Until now, neighbor-embedding-based (NE) algorithms for super-resolution (SR) have carried out two independent processes to synthesize high-resolution (HR) image patches. In the first process, neighbor search is performed using the Euclidean distance metric, and in the second process, the optimal weights are determined by solving a constrained least squares problem. However, the separate processes are not optimal. In this paper, we propose a sparse neighbor selection scheme for SR reconstruction. We first predetermine a larger number of neighbors as potential candidates and develop an extended Robust-SL0 algorithm to simultaneously find the neighbors and to solve the reconstruction weights. Recognizing that the k-nearest neighbor (k-NN) for reconstruction should have similar local geometric structures based on clustering, we employ a local statistical feature, namely histograms of oriented gradients (HoG) of low-resolution (LR) image patches, to perform such clustering. By conveying local structural information of HoG in the synthesis stage, the k-NN of each LR input patch is adaptively chosen from their associated subset, which significantly improves the speed of synthesizing the HR image while preserving the quality of reconstruction. Experimental results suggest that the proposed method can achieve competitive SR quality compared with other state-of-the-art baselines.

Illumination Invariant Face Recognition Using the Statistical Features of BDIP and Wavelet Transform

In this paper, an efficient feature extraction method based on local statistics features of block difference of inverse probabilities (BDIP) and the wavelet transform is proposed for face recognition. In the proposed method, the BDIPs are first computed in a face in order to overcome the variation of illumination and facial expressions. The obtained BDIP image is then decomposed into wavelet subbands. In order to reduce the dimensionality of the feature vector, each BDIP subband is partitioned into a set of blocks. The means and variances are then calculated from all the blocks in each subband and are fused into a feature vector. Experimental results on ORL and FERET databases show that the proposed method achieves higher recognition accuracies than the wavelet-based methods with higher dimensionality reduction of the feature vector. It also outperforms the other well known methods such as PCA and the DCT with the zigzag scanning. Index Terms—Face recognition (FR), discrete wavelet transforms (DWT), wavelet packet decomposition (WPD), block difference of inverse probabilities (BDIP), support vector machine (SVM).

Scene Change Detection Algorithm for H.264 using Dynamic Threshold and Fast Motion Estimation Algorithm based on PSO

Most of the existed scene change detection algorithms use fixed thresholds for finding the scene change. These thresholds are obtained by empirically or they must be calculated before the detection after the whole sequence is obtained. If Videos having high scene complexity and variation, then the Performance of most scene change algorithms decreases considerably. In this paper, we study the correlation between local statistical characteristics, scene duration and scene change. Based on this analysis, we further propose and implement a scene change algorithm for H.264 codec, defining an automated, dynamic threshold model with fast motion estimation algorithm with low complexity which can efficiently trace scene changes.

Spatially Adapted Total Variation Model to Remove Multiplicative Noise

Multiplicative noise removal based on total variation (TV) regularization has been widely researched in image science. In this paper, inspired by the spatially adapted methods for denoising Gaussian noise, we develop a variational model, which combines the TV regularizer with local constraints. It is also related to a TV model with spatially adapted regularization parameters. The automated selection of the regularization parameters is based on the local statistical characteristics of some random variable. The corresponding subproblem can be efficiently solved by the augmented Lagrangian method. Numerical examples demonstrate that the proposed algorithm is able to preserve small image details, whereas the noise in the homogeneous regions is sufficiently removed. As a consequence, our method yields better denoised results than those of the current state-of-the-art methods with respect to the signal-to-noise-ratio values.

An improved region-based model with local statistical features for image segmentation

In this paper, we propose a new region-based active contour model (ACM) for image segmentation. In particular, this model utilizes an improved region fitting term to partition the regions of interests in images depending on the local statistics regarding the intensity and the magnitude of gradient in the neighborhood of a contour. By this improved region fitting term, images with noise, intensity non-uniformity, and low-contrast boundaries can be well segmented. Integrated with the duality theory and the anisotropic diffusion process based on structure tensor, a new regularization term is defined through the duality formulation to penalize the length of active contour. By this new regularization term, the structural information of images is utilized to improve the ability of capturing the geometric features such as corners and cusps. From a numerical point of view, we minimize the energy function of our model by an efficient dual algorithm, which avoids the instability and the non-differentiability of traditional numerical solutions, e.g. the gradient descent method. Experiments on medical and natural images demonstrate the advantages of the proposed model over other segmentation models in terms of both efficiency and accuracy.

Structural visualization of sequential DNA data

To date, comparing and visualizing genome sequences remain challenging due to the large genome size. Existing approaches take advantage of the stable property of oligonucleotides and exhibit the main characteristics of the whole genome, yet they commonly fail to show progression patterns of the genome adjustably. This paper presents a novel visual encoding technique, which not only supports the binning process (phylogenetic analysis), but also allows the sequential analysis of the genome. The key idea is to regard the combination of each k-nucleotide and its reverse complement as a visual word, and to represent a long genome sequence with a list of local statistical feature vectors derived from the local frequency of the visual words. Experimental results on a variety of examples demonstrate that the presented approach has the ability to quickly and intuitively visualize DNA sequences, and to help the user identify regions of differences among multiple datasets.

일강우 내삽을 이용한 일유량 시뮬레이션 및 앙상블 유량 발생

입력자료의 불확실성은 강우-유출 모의에서 중요한 불확실성 요소 중의 하나이다. 본 연구에서는 먼저 세 가지의 서로 다른 내삽 기법을 통해 계산된 강수 입력 자료 (관측값을 각 소유역의 중심점으로 내삽하여 추정한 입력자료임)들이 강우-유출 모형에 미치는 영향을 분포형 수문모형 (PRMS)을 이용하여 분석하였으며, 내삽오차를 바탕으로 발생한 입력자료를 앙상블 유량 예측에 이용하는 과정을 수문학적으로 서로 다른 두개 하천 유역에 적용하였다. 또한 Monte Carlo기법을 이용하여 수문 모형의 매개변수가 서로 다른 입력자료의 특성에 따라 변화하는 양상을 구분하여 보았다. 본 연구에서 제시된 앙상블 유량 예측방법은 기상 예측 및 기상 모형의 결과물 등의 입력자료를 이용함으로써 중/장기 유량 예측에 활용될 수 있을 것으로 판단된다. Input uncertainty is one of the major sources of uncertainty in hydrologic modeling. In this paper, first, three alternate rainfall inputs generated by different interpolation schemes were used to see the impact on a distributed watershed model. Later, the residuals of precipitation interpolations were tested as a source of ensemble streamflow generation in two river basins in the U.S. Using the Monte Carlo parameter search, the relationship between input and parameter uncertainty was also categorized to see sensitivity of the parameters to input differences. This analysis is useful not only to find the parameters that need more attention but also to transfer parameters calibrated for station measurement to the simulation using different inputs such as downscaled data from weather generator outputs. Input ensembles that preserves local statistical characteristics are used to generate streamflow ensembles hindcast, and showed that the ensemble sets are capturing the observed steamflow properly. This procedure is especially important to consider input uncertainties in the simulation of streamflow forecast.

Benthic Classifications Using Bathymetric LIDAR Waveforms and Integration of Local Spatial Statistics and Textural Features

The scope of this research is to assess benthoscape discrimination by airborne light detection and ranging (LIDAR) bathymetry (ALB) on the basis of statistical parameters derived from the LIDAR waveforms, textural information, and local spatial statistics. Analysis of the underwater camera stations allowed clustering of the stations into groups on the basis of their habitat composition (β-diversity). Twelve descriptive statistics describing the shape of the bottom part of the waveform, also called 12 benthic parameters, were used for discriminating four benthic habitats. A K-means classification and a supervised method based on the Support Vector Machine (SVM) were applied to this dataset, and overall accuracies of 67.7% and 89.9% were obtained, respectively. Geostatistical analyses, using 11 textural measures, defined by the gray-level occurrence matrix (GLOM) and the gray-level co-occurrence matrix (GLCM), and three local spatial statistics were then applied to the 12 benthic parameters to enhance the SVM classification performance. The assessment of the contribution of geostatistics into benthic class segmentation was achieved by computation of separability distance. Mean (from the GLOM), mean (from the GLCM) and the local Getis-Ord statistic yielded the best rates of discrimination. These added metrics, integrated with bands related to the 12 benthic parameters, showed that the rate of correct (supervised) classification was thereby improved and increased by 5.3%. Finally, the first four principal components (PCs) (i.e., 90.41% of the 12 parameter variances, boosted by the three best geostatistics) brought out an overall accuracy of 93.3%, showing evidence for optimizing the classification processing.

A Rapid Automatic Image Registration Method Based on Improved SIFT

Abstract This paper presents an image registration method based on the improved SIFT feature extraction algorithm, which takes local statistical features and corner feature into account. With this improved algorithm, an automatic registration of narrow- baseline images can be achieved which can greatly improve the speed and accuracy of the registration.

Fast algorithm for maneuvering target detection in SAR imagery based on gridding and fusion of texture features

Designing detection algorithms with high efficiency for Synthetic Aperture Radar (SAR) imagery is essential for the operator SAR Automatic Target Recognition (ATR) system. This work abandons the detection strategy of visiting every pixel in SAR imagery as done in many traditional detection algorithms, and introduces the gridding and fusion idea of different texture features to realize fast target detection. It first grids the original SAR imagery, yielding a set of grids to be classified into clutter grids and target grids, and then calculates the texture features in each grid. By fusing the calculation results, the target grids containing potential maneuvering targets are determined. The dual threshold segmentation technique is imposed on target grids to obtain the regions of interest. The fused texture features, including local statistics features and Gray-Level Co-occurrence Matrix (GLCM), are investigated. The efficiency and superiority of our proposed algorithm were tested and verified by comparing with existing fast detection algorithms using real SAR data. The results obtained from the experiments indicate the promising practical application value of our study.

A Multi-Scale Vectorial L τ -TV Framework for Color Image Restoration

A general multi-scale vectorial total variation model with spatially adapted regularization parameter for color image restoration is introduced in this paper. This total variation model contains an Lτ-data fidelity for any τC[1,2]. The use of a spatial dependent regularization parameter improves the reconstruction of features in the image as well as an adequate smoothing for the homogeneous parts. The automated adaptation of this regularization parameter is made according to local statistical characteristics of the noise which contaminates the image. The corresponding multiscale vectorial total variation model is solved by Fenchel-duality and inexact semismooth Newton techniques. Numerical results are presented for the cases τ=1 and τ=2 which reconstruct images contaminated with salt-and-pepper noise and Gaussian noise, respectively.

Masking of Time-Frequency Patterns in Applications of Passive Underwater Target Detection

Spectrogram analysis of acoustical sounds for underwater target classification is utilized when loud nonstationary interference sources overlap with a signal of interest in time but can be separated in time-frequency ( ) domain. We propose a signal masking method which in a plane combines local statistical and morphological features of the signal of interest. A dissimilarity measure of adjacent cells is used for local estimation of entropy , followed by estimation of entropy difference, where is calculated along the time axis at a mean frequency and is calculated along the frequency axis at a mean time of the window, respectively. Due to a limited number of points used in estimation, the number of possible values, which define a primary mask, is also limited. secondary mask is defined using morphological operators applied to, for example, and . We demonstrate how primary and secondary masks can be used for signal detection and discrimination, respectively. We also show that the proposed approach can be generalized within the framework of Genetic Programming.

Locating and parsing bibliographic references in HTML medical articles

The set of references that typically appear toward the end of journal articles is sometimes, though not always, a field in bibliographic (citation) databases. But even if references do not constitute such a field, they can be useful as a preprocessing step in the automated extraction of other bibliographic data from articles, as well as in computer-assisted indexing of articles. Automation in data extraction and indexing to minimize human labor is key to the affordable creation and maintenance of large bibliographic databases. Extracting the components of references, such as author names, article title, journal name, publication date and other entities, is therefore a valuable and sometimes necessary task. This paper describes a two-step process using statistical machine learning algorithms, to first locate the references in HTML medical articles and then to parse them. Reference locating identifies the reference section in an article and then decomposes it into individual references. We formulate this step as a two-class classification problem based on text and geometric features. An evaluation conducted on 500 articles drawn from 100 medical journals achieves near-perfect precision and recall rates for locating references. Reference parsing identifies the components of each reference. For this second step, we implement and compare two algorithms. One relies on sequence statistics and trains a Conditional Random Field. The other focuses on local feature statistics and trains a Support Vector Machine to classify each individual word, followed by a search algorithm that systematically corrects low confidence labels if the label sequence violates a set of predefined rules. The overall performance of these two reference-parsing algorithms is about the same: above 99% accuracy at the word level, and over 97% accuracy at the chunk level.

A more secure steganography based on adaptive pixel-value differencing scheme

Pixel-value differencing (PVD) based steganography is one of popular approaches for secret data hiding in the spatial domain. However, based on extensive experiments, we find that some statistical artifacts will be inevitably introduced even with a low embedding capacity in most existing PVD-based algorithms. In this paper, we first analyze the common limitations of the original PVD and its modified versions, and then propose a more secure steganography based on a content adaptive scheme. In our method, a cover image is first partitioned into small squares. Each square is then rotated by a random degree of 0, 90, 180 or 270. The resulting image is then divided into non-overlapping embedding units with three consecutive pixels, and the middle one is used for data embedding. The number of embedded bits is dependent on the differences among the three pixels. To preserve the local statistical features, the sort order of the three pixel values will remain the same after data hiding. Furthermore, the new method can first use sharper edge regions for data hiding adaptively, while preserving other smoother regions by adjusting a parameter. The experimental results evaluated on a large image database show that our method achieves much better security compared with the previous PVD-based methods.

Dense Sampling Low-Level Statistics of Local Features

Generic image recognition techniques are widely studied for automatic image indexing. However, many of these methods are computationally too heavy for a practically large setup. Thus, for realizing scalability, it is important to properly balance the trade-off between performance and computational cost. In recent years, methods based on a bag-of-keypoints approach have been successful and widely used. However, the preprocessing cost for building visual words becomes immense in large-scale datasets. On the other hand, methods based on global image features have been used for a long time. Because global image features can be extracted rapidly, it is relatively easy to use them with large datasets. However, the performance of global feature methods is usually poor compared to the bag-of-keypoints methods. This paper proposes a simple but powerful scheme of boosting the performance of global image features by densely sampling low-level statistical moments of local features. Also, we use a scalable learning and classification method which is substantially lighter than a SVM. Our method achieved performance comparable to state-of-the-art methods despite its remarkable simplicity.

Robust Source Localization in Reverberant Environments Based on Weighted Fuzzy Clustering

Successful localization of sound sources in reverberant enclosures is an important prerequisite for many spatial signal processing algorithms. We investigate the use of a weighted fuzzy <b xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">c</b> -means cluster algorithm for robust source localization using location cues extracted from a microphone array. In order to increase the algorithm's robustness against sound reflections, we incorporate observation weights to emphasize reliable cues over unreliable ones. The weights are computed from local feature statistics around sound onsets because it is known that these regions are least affected by reverberation. Experimental results illustrate the superiority of the method when compared with standard fuzzy clustering. The proposed algorithm successfully located two speech sources for a range of angular separations in room environments with reverberation times of up to 600 ms.

Facial expression recognition based on Local Binary Patterns: A comprehensive study

Automatic facial expression analysis is an interesting and challenging problem, and impacts important applications in many areas such as human–computer interaction and data-driven animation. Deriving an effective facial representation from original face images is a vital step for successful facial expression recognition. In this paper, we empirically evaluate facial representation based on statistical local features, Local Binary Patterns, for person-independent facial expression recognition. Different machine learning methods are systematically examined on several databases. Extensive experiments illustrate that LBP features are effective and efficient for facial expression recognition. We further formulate Boosted-LBP to extract the most discriminant LBP features, and the best recognition performance is obtained by using Support Vector Machine classifiers with Boosted-LBP features. Moreover, we investigate LBP features for low-resolution facial expression recognition, which is a critical problem but seldom addressed in the existing work. We observe in our experiments that LBP features perform stably and robustly over a useful range of low resolutions of face images, and yield promising performance in compressed low-resolution video sequences captured in real-world environments.

Fusion of Local Statistical Parameters for Buried Underwater Mine Detection in Sonar Imaging

Detection of buried underwater objects, and especially mines, is a current crucial strategic task. Images provided by sonar systems allowing to penetrate in the sea floor, such as the synthetic aperture sonars (SASs), are of great interest for the detection and classification of such objects. However, the signal-to-noise ratio is fairly low and advanced information processing is required for a correct and reliable detection of the echoes generated by the objects. The detection method proposed in this paper is based on a data-fusion architecture using the belief theory. The input data of this architecture are local statistical characteristics extracted from SAS data corresponding to the first-, second-, third-, and fourth-order statistical properties of the sonar images, respectively. The interest of these parameters is derived from a statistical model of the sonar data. Numerical criteria are also proposed to estimate the detection performances and to validate the method.

A machine vision inspector for beer bottle

A machine-vision-based beer bottle inspector is presented. The mechanical structure and electric control system are illustrated in detail. A method based on the histogram of edge points is applied for real-time determination of inspection area. For defect detection of bottle wall and bottle bottom, an algorithm based on local statistical characteristics is proposed. In bottle finish inspection, two artificial neural networks are used for low-level inspection and high-level judgment, respectively. A prototype was developed and experimental results demonstrate the feasibility of the inspector. Inspections performed by the prototype have proved the effectiveness and value of proposed algorithms in automatic real-time inspection.