The Infona portal uses cookies, i.e. strings of text saved by a browser on the user's device. The portal can access those files and use them to remember the user's data, such as their chosen settings (screen view, interface language, etc.), or their login data. By using the Infona portal the user accepts automatic saving and using this information for portal operation purposes. More information on the subject can be found in the Privacy Policy and Terms of Service. By closing this window the user confirms that they have read the information on cookie usage, and they accept the privacy policy and the way cookies are used by the portal. You can change the cookie settings in your browser.
This paper presents a fast, accurate and novel approach for the problem of flow segmentation in dense and very dense moving crowds. First, optical flow method is used to remove background noise of scenes. Second, the angle information of foreground velocity field is turned into gray level image and histogram curve is employed to find out extreme points. Finally, the updated minimum points can be utilized...
Finger vein recognition has been identified as a stable biometrics technique that has many advantages as compared to other techniques. The biggest challenge that is faced while using this technique is to make the features rotarionally invariant. In this paper, local binary pattern variance (LBPV) is proposed to address this challenge and to characterize the local contrast information into the one-dimensional...
This paper presents a robust multi-view vehicle detection based on the Histogram of Oriented Gradient (HOG)-Histograms of Census Transform (HCT) features and the mixtures of deformable part models. As some virtual features of vehicle in single view, such as headlight, taillight and edges can not been directly used, we develop a new HOG-HCT feature to describe the vehicle structure feature in multi-view...
In machine vision inspection with area array camera, inspecting windows placement needs to be optimized. It aims to minimize the number of windows while covering whole inspected objects. A knowledge inductive search approach is proposed in this work to randomly search optimal window positions with partly prior knowledge direction and partly posteriori knowledge induction. Experimental results show...
Using Behavior Entropy model, we introduce a novel method to detect and localize abnormal behaviors in crowd scenes. Our key insight is to estimate the behavior entropy of each pixel and whole scene by considering defined pixels' behavior certainty. For this purpose, we introduce information theory and energetics concept to define pixel's behavior certainty based on video's spatial-temporal information...
Feature fusion can effectively improve the steganographic detection capability, but the previous researches of feature fusion in JPEG image steganography detection rarely considered the nonlinear correlation of features. This paper analyzes the correlation of JPEG image steganographic features and fuses features with lowest correlation to obtain better detection capability based on KCCA (Kernel canonical...
Pedestrian detection is of much importance for its practical applications. This paper develops a novel pedestrian detection system which consists of three stages: motion region detection based on background modeling, feature extraction in the guidance of prior information, and map-based classification applying support vector machine (SVM) and Adaboost. First of all, an adaptive Gaussian Mixture Model...
We present a novel method for object tracking using global and local states of object in video surveillance application. Most traditional object models using global appearance cannot handle partial occlusion effectively. The unoccluded part of partially visible object retains invariable appearance. Therefore, we introduce global and local dynamics model as our object model to overcome partial occlusion...
License Plate Recognition (LPR) plays an important role on the traffic monitoring and parking management. In this paper, an updated algorithm is applied into the vehicle license plate identification, which is mainly based on derived kernel through visual cortex. With the two-layer derived kernel on neural response and first nearest classification method applied to character and numeral recognition,...
Image saliency attempts to describe the most conspicuous part in an input image by mimicking human visual selective attention mechanism. Naturally, it could be adopted for improving object recognition. To demonstrate the effectiveness of saliency in object recognition, this paper proposes a salient hierarchical model. First, the traditional saliency model is modified for more robust saliency estimation...
Fault location estimation, which can actually be addressed as a classification or categorization problem, is a vital feature in protective relaying scheme for power transmission lines. In this paper, a novel three-phase fault location approach based on a Multiple Classifier System (MCS) not only using the information of the faulty lines but also considering others lines adjacent with the faulty line...
In this paper, Gabor frame, which is generated by multiple generators of L2(Rd), is studied. Firstly, several sufficient conditions and necessary conditions are presented for multi-Gabor frame (MGF). Then, these results are extended to the matrices which satisfy A = (BT)−1, and two corollaries are given for Gabor frame {EmBTnAgl: m, n ∊ Zd; l = 1, …, L}. Finally, an example is provided for illustration.
Empirical mode decomposition (EMD) is a data driven processing algorithm, which has no predetermined filter. It is able to perfectly analyze the nonlinear and nonstationary signals. In EMD decomposition processing, the envelopes are computed by spline interpolation, which is time-consuming. In this work, firstly, we proposed a boundary extending method based on linear prediction and boundary extreme...
Based on the characteristics of the pulse singularity signals, a new method for detecting the pulse singularity signals is proposed using the Bernoulli probability to describe singularity points. The proposed method converts the detection problem to a parameter estimation problem and solves it using a maximum likelihood method. The experimental results show that the proposed method identifies pulse...
A quantitative calibration model for Infrared spectroscopy using continuous wavelet transform combined with genetic algorithm is presented in this paper. We propose three scale selection methods in continuous wavelet transform and provide the comparisons with the general preprocessing methods. Experimental results show that selectively combining scales results in a quantitative model with better performance...
Traditional Fourier analysis is applied to signal processing, but it often causes ‘jump’ at the edges. In this paper, a quantitative analysis of the ‘jump’ and a wavelet series construction method by folding and integral operator in the H1[0,1] space are introduced. Its good properties are discussed here for the first time. Without the need of pre-filtering and boundary extension, the proposed method...
In this paper, we study minimum-energy multi-wavelets frames Ψ = {ψ1, ψ2, …, ψN} with dilation factor 3 for L2(R), Ψ correspond to some refutable functions with compact support. Firstly, the concept of minimum-energy multi-wavelets frames is generalized to dilation factor 3 and a precise characterization of Ψ is given in terms of the Laurent matrix polynomial of the refinable functions. Secondly,...
In relation to the problems of time variability and the complexity of mine earthquake signal while blasting mining, a kind of wavelet soft threshold filter method is proposed. In this paper, the signal before mine earthquake wave arrival time is considered as noise, and is used to determine the wavelet threshold. Then Db4 wavelet is chosen as the wavelet basis function for the wavelet soft threshold...
Human binaural processing might enhance signal sounds in noisy environments. Binaural speech enhancement with two outputs facilitates merits of both signal processing itself and that by human binaural processing. Most previous studies in this area have implemented signal processing in the time and frequency domains. The use of wavelet transform (WT) appears to be promising because it has a scale domain...
A novel signal processing method, called sensitive frequency band (SFB) selection, is proposed based on the correlation coefficients between the original signal and its reconstructed approximation and detail parts. By identifying the noised-related and fault-related SFBs, the SFB selection method can be used to reduce noise and extract fault feature effectively for a signal. This method is applied...
Set the date range to filter the displayed results. You can set a starting date, ending date or both. You can enter the dates manually or choose them from the calendar.