Learning Dual-Domain Calibration and Distance-Driven Correlation Filter:A Probabilistic Perspective for UAV Tracking

Unmanned Aerial Vehicle(UAV)tracking has been possible because of the growth of intelligent information technology in smart cities,making it simple to gather data at any time by dynamically monitoring events,people,the environment,and other aspects in the city.The traditional filter creates a model to address the boundary effect and time filter degradation issues in UAV tracking operations.But these methods ignore the loss of data integrity terms since they are overly dependent on numerous explicit previous regularization terms.In light of the aforementioned issues,this work suggests a dual-domain Jensen-Shannon divergence correlation filter(DJSCF)model address the probability-based distance measuring issue in the event of filter degradation.The two-domain weighting matrix and JS divergence constraint are combined to lessen the impact of sample imbalance and distortion.Two new tracking models that are based on the perspectives of the actual probability filter distribution and observation probability filter distribution are proposed to translate the statistical distance in the online tracking model into response fitting.The model is roughly transformed into a linear equality constraint issue in the iterative solution,which is then solved by the alternate direction multiplier method(ADMM).The usefulness and superiority of the suggested strategy have been shown by a vast number of experimental findings.

Collaborative Filtering Algorithms Based on Kendall Correlation in Recommender Systems

In this work, Kendall correlation based collaborative filtering algorithms for the recommender systems are proposed. The Kendall correlation method is used to measure the correlation amongst users by means of considering the relative order of the users＇ ratings. Kendall based algorithm is based upon a more general model and thus could be more widely applied in e-commerce. Another discovery of this work is that the consideration of only positive correlated neighbors in prediction, in both Pearson and Kendall algorithms, achieves higher accuracy than the consideration of all neighbors, with only a small loss of coverage.

NLoS Mitigation in ToA Localization Based on Spatial Correlation Filter and Iterative Minimum Residual

To mitigate the Non-Line-of-Sight (NLoS) error which seriously affects the localization accuracy and robustness in complex indoor environment,a novel Iterative Minimum Residual (IMR) based on the consistency hypothesis of the residual and the error is proposed in this paper.It chooses the best subset of measurements to calculate the coordinates of the unknown node by comparing the residuals obtained with different subsets of beacons.To reduce the time complexity of the IMR algorithm,Spatial Correlation Filter (SCF) is also proposed,which can remove the most serious NLoS distance with low calculation cost.Combined with the proposed SCF and IMR algorithm,nodes can be localized with high accuracy and low time complexity.Experimental results with real dataset demonstrate that the proposed algorithm can identify the NLoS range effectively with about 50% time cost of employing SCF only.

A Method for Incipient Fault Diagnosis of Roller Bearings Based on the Wavelet Transform Correlation Filter and Hilbert Transform

Noise is the biggest obstacle that makes the incipient fault diagnosis results of roller bearings uncorrected; a new method for diagnosing incipient fault of roller bearings based on the Wavelet Transform Correlation Filter and Hilbert Transform was proposed. First, the weak fault information features are picked up from the roller bearings fault vibration signals by use of a de-noising characteristic of the Wavelet Transform Correlation Filter as the preprocessing of the Hilbert Envelope Analysis. Then, in order to get fault features frequency, de-noised wavelet coefficients of high scales which represent high frequency signal were analyzed by Hilbert Envelope Spectrum Analysis. The simulation signals and diagnosing examples analysis results reveal that the proposed method is more effective than the method of direct wavelet coefficients-Hilbert Transform in de-noising and clarifying roller bearing incipient fault.

Harmonic detection an AC excited generation system based on in-phase correlation filtering

The paper reports results of investigation on the harmonic detection technique of a complicated power supply system such as an AC excited generation system, which has a variable fundamental frequency and low order harmonics with rich sub-harmonics whose frequencies are lower than the fundamental one. The in-phase correlation filtering technique, based on the frequency shifting principle, is proposed in this paper.Theoretical analysis and experimental results validate the effectiveness of this technique for the harmonic detections of AC excited generation systems.

An Adaptive Padding Correlation Filter With Group Feature Fusion for Robust Visual Tracking

In recent visual tracking research,correlation filter(CF)based trackers become popular because of their high speed and considerable accuracy.Previous methods mainly work on the extension of features and the solution of the boundary effect to learn a better correlation filter.However,the related studies are insufficient.By exploring the potential of trackers in these two aspects,a novel adaptive padding correlation filter(APCF)with feature group fusion is proposed for robust visual tracking in this paper based on the popular context-aware tracking framework.In the tracker,three feature groups are fused by use of the weighted sum of the normalized response maps,to alleviate the risk of drift caused by the extreme change of single feature.Moreover,to improve the adaptive ability of padding for the filter training of different object shapes,the best padding is selected from the preset pool according to tracking precision over the whole video,where tracking precision is predicted according to the prediction model trained by use of the sequence features of the first several frames.The sequence features include three traditional features and eight newly constructed features.Extensive experiments demonstrate that the proposed tracker is superior to most state-of-the-art correlation filter based trackers and has a stable improvement compared to the basic trackers.

Heteronuclear-filtered ^(1)H homonuclear multi-quantum correlation experiment at 100 kHz magic-angle spinning Dedicated to Professor Chaohui Ye on the occasion of his 80th birthday

Remarkable advances in fast magic-angle spinning(MAS)techniques significantly improve the resolution of^(1)H solid-state nuclear magnetic resonance(NMR)spectra.Here,we introduce a heteronuclear-filtered^(1)H homonuclear multi-quantum(MQ)correlation strategy available at a MAS rate of 100 kHz by combining^(1)H{X}heteronuclear-filtered methods and^(1)H homonuclear MQ correlation experiments.The proposed strategy was applied to selectively extract^(1)H signals of aluminum lactate(Al-Lac)in a mixture of Al-Lac and zinc lactate(Zn-Lac)using 27Al-filtered methods(i.e.,^(1)H{27Al}heteronuclear multiple quantum correlation(HMQC)or^(1)H{27Al}symmetry-based resonance-echo saturationpulse double-resonance(S-RESPDOR)).We demonstrate that incorporating these 27Al-filtered methods into two-dimensional(2D)^(1)He^(1)H double-quantum(DQ)/single-quantum(SQ),triple-quantum(TQ)/SQ,and even three-dimensional(3D)27Al/^(1)H(DQ)/^(1)H(SQ)experiments can facilitate the acquisition of spectra without signal overlap and targeted characterization of the^(1)H species surrounding 27Al sites.The proposed strategy is considered to efficiently extract key structural information from complex spin systems.

A Collaborative Filtering Recommendation Algorithm Based on the Difference and the Correlation of Users’Ratings

The traditional similarity algorithm in collaborative filtering mainly pay attention to the similarity or correlation of users’ratings,lacking the consideration of difference of users’ratings.In this paper,we divide the relationship of users’ratings into differential part and correlated part,proposing a similarity measurement based on the difference and the correlation of users’ratings which performs well with non-sparse dataset.In order to solve the problem that the algorithm is not accurate in spare dataset,we improve it by prefilling the vacancy of rating matrix.Experiment results show that this algorithm improves significantly the accuracy of the recommendation after prefilling the rating matrix.

Kernelized Correlation Filter Target Tracking Algorithm Based on Saliency Feature Selection

To address the problem of using fixed feature and single apparent model which is difficult to adapt to the complex scenarios, a Kernelized correlation filter target tracking algorithm based on online saliency feature selection and fusion is proposed. It combined the correlation filter tracking framework and the salient feature model of the target. In the tracking process, the maximum Kernel correlation filter response values of different feature models were calculated respectively, and the response weights were dynamically set according to the saliency of different features. According to the filter response value, the final target position was obtained, which improves the target positioning accuracy. The target model was dynamically updated in an online manner based on the feature saliency measurement results. The experimental results show that the proposed method can effectively utilize the distinctive feature fusion to improve the tracking effect in complex environments.

Most relevant weighted filtering with one-step singular correlation recognition

Based on the recognition of one-step singular correlation and the remedying methods obtained before,the correlation properties of the neighborhood pixels and the characteristics of image de-noising were analyzed.A kind of most relevant weighted filtering method based on one-step singular correlation recognition(OSSC-MRWF)was put forward.The simulation experiments were done and the comparison with some commonly used methods under salt-and-pepper noises was made.The results show that the proposed method can not only effectively recognize salt-and-pepper noises and mend up the noise points,but also protect the original information such as the edge details very well.The accuracy and performance indicators are further improved considerably.

Signal Processing Circuit Design of Infrared Detection System with SO2 Concentration Based on Correlation Filter Technology

Signals from infrared detector are very weak in SO2 concentration measuring system.In order to improve the sensitivity of detection,combining with filter correlation technology and infrared absorption principle,the weak signal processing circuit is designed according to correlation detection technology.Under laboratory conditions,system performance of SO2 concentration is tested,and the experimental data are analyzed and processed.Then relationship of SO2 concentration and the measuring voltage is provided to prove that the design improves measuring sensitivity of the system.

New sigma point filtering algorithms for nonlinear stochastic systems with correlated noises

New sigma point filtering algorithms,including the unscented Kalman filter(UKF) and the divided difference filter(DDF),are designed to solve the nonlinear filtering problem under the condition of correlated noises.Based on the minimum mean square error estimation theory,the nonlinear optimal predictive and correction recursive formulas under the hypothesis that the input noise is correlated with the measurement noise are derived and can be described in a unified framework.Then,UKF and DDF with correlated noises are proposed on the basis of approximation of the posterior mean and covariance in the unified framework by using unscented transformation and second order Stirling's interpolation.The proposed UKF and DDF with correlated noises break through the limitation that input noise and measurement noise must be assumed to be uncorrelated in standard UKF and DDF.Two simulation examples show the effectiveness and feasibility of new algorithms for dealing with nonlinear filtering issue with correlated noises.

Biometric feature extraction using local fractal auto-correlation

Image texture feature extraction is a classical means for biometric recognition. To extract effective texture feature for matching, we utilize local fractal auto-correlation to construct an effective image texture descriptor. Three main steps are involved in the proposed scheme： （i） using two-dimensional Gabor filter to extract the texture features of biometric images; （ii） calculating the local fractal dimension of Gabor feature under different orientations and scales using fractal auto-correlation algorithm; and （iii） linking the local fractal dimension of Gabor feature under different orientations and scales into a big vector for matching. Experiments and analyses show our proposed scheme is an efficient biometric feature extraction approach.

Generalized Correlativity of Median Filtering Operator on Signals

The generalized correlativity of input signal and output signal of a stack filtering operator is defined and used for numerously measuring these filtering operators's behavior in removing noise in signals. We show that under the criterion of the generalized correlativity, of stack filtering operators the median filtering operator is optimal, which implies that this filtering operator possesses better filtering behavior than the others.

A correlation OPTS algorithm for reducing peak to average power ratio of FBMC-OQAM systems

A correlation overlapping partial transmit sequence(C-OPTS) algorithm is proposed to solve the issue of high complexity of overlapping partial transmit sequence(OPTS) algorithm in suppressing the peak to average power ratio(PAPR) of filter bank multicarrier-offset quadrature amplitude modulation(FBMC-OQAM) signals.The V subblocks in partial transmit sequence(PTS) are regrouped into U combinations according to the correlation coefficient p,and overlapping subblocks are allowed between adjacent groups.The search starts from the first group and sets the phase factors of the subsequent groups to 1.When the phase factors of the non-overlapping subblocks in the first group are determined,the subsequent groups are searched in turn to determine their respective phase factors.Starting from the second data block,the data overlapped with it should be taken into account when determining its optimal phase factor vector.Theoretical analysis and simulation results indicate that compared with the OPTS algorithm,the proposed algorithm can significantly reduce the computational complexity at the cost of slight deterioration of PAPR performance.Meanwhile,compared with the even-odd iterative double-layers OPTS(ID-OPTS) algorithm,it can further reduce the complexity and obtain a better PAPR suppression effect.

SELF-TUNING MEASUREMENT FUSION KALMAN FILTER WITH CORRELATED MEASUREMENT NOISES

For the multisensor system with correlated measurement noises and unknown noise statistics, based on the solution of the matrix equations for correlation function, the on-line estimators of the noise variances and cross-covariances is obtained. Further, a self-tuning weighted measurement fusion Kalman filter is presented, based on the Riccati equation. By the Dynamic Error System Analysis (DESA) method, it rigorously proved that the presented self-tuning weighted measurement fusion Kalman filter converges to the optimal weighted measurement fusion steady-state Kalman filter in a realization or with probability one, so that it has asymptotic global optimality. A simulation example for a target tracking system with 3-sensor shows that the presented self-tuning measurement fusion Kalman fuser converges to the optimal steady-state measurement fusion Kalman fuser.

Digital Cross-Correlation Detection of Multi-Laser Beams Measuring System for Wind Field Detection

A cross-correlation detection method to process backscatter signals of multi-laser beams measuring （MLBM） is presented, which can be firstly filtered by the digital filter composed of average median filter and finite impulse response （FIR） digital filter. The processing of backscatter signals using single-pulse and three-pulse cross-correlation detection methods is depicted in detail. From calculation results, the multi-pulse cross-correlation detection could effectively improve signal-to-noise ratio （SNR）. Finally, both wind velocity and direction are determined by the peak-delay method based on the correlation function which shows high measuring precision and high SNR of the MLBM system with the assistance of the digital cross- correlation detection.

采用局部-全局区域重检测机制的无人机长期跟踪算法

为解决基础跟踪器面对遮挡和移出视野等长期跟踪场景时易出现跟踪失败等问题,提出了一种基于局部-全局区域重检测的无人机长期跟踪算法。设计了基础滤波器,将高置信度样本与其结合,并融入自适应时空正则化,解决了滤波器退化问题,提高了模型鲁棒性以及复杂场景下的性能;优化了滤波器更新策略,通过评价跟踪结果进行自适应更新;设计快速尺度滤波器,解决了跟踪过程中的尺度变化问题;设计了局部-全局区域重检测机制,跟踪失败时启动重检测器恢复跟踪目标,先完成局部区域重检测,若恢复跟踪失败,再利用全局区域重检测器继续恢复目标跟踪状态。实验结果表明:所提算法在UAV20L数据集上的精确度和准确率分别可达0.724和0.621,与基于时空正则化相关滤波器的跟踪算法(STRCF)相比分别提升了25.9%和20.6%,与同类主流算法相比,跟踪效果得到提升,证明了算法的有效性。

飞蛾扑火优化的尺度比例感知空间长期跟踪器

针对无人机长期跟踪过程中尺度变换导致目标丢失和跟踪精度低的问题,提出了一种基于飞蛾扑火优化(moth-flame optimization,MFO)的尺度比例感知空间长期跟踪器。首先,设计了高斯初始化以代替飞蛾扑火优化算法的随机初始化策略,降低优化算法在跟踪过程中的计算复杂度,减少算力浪费;其次,结合快速梯度直方图特征,构建了改进的飞蛾扑火优化跟踪器;然后,为了解决无人机航拍长期跟踪中目标尺度变化的问题,设计了一种自适应尺度变换的判别尺度空间跟踪(discriminative scale space tracking,DSST)算法,进一步提出了一种尺度比例感知空间跟踪器,解决了尺度滤波器中因长宽比固定而导致的跟踪漂移;同时,分析了滤波器响应峰值在各背景下的变化情况,提出了一种能反映环境变化下跟踪置信度的指标,并通过置信度将MFO优化跟踪框架与尺度比例感知空间跟踪器相结合,解决了尺度变化与长期跟踪目标丢失的问题;最后,在无人机长期跟踪数据集上开展了性能验证。结果表明:提出的算法可有效防止漂移现象的发生,提升跟踪效率;与目前跟踪领域中12种同类文献算法进行对比可知,提出的算法精度较高,满足实时性,能够有效解决无人机长期跟踪下的尺度变化及目标丢失等问题。

基于核相关滤波和卡尔曼滤波预测的混合跟踪方法

针对核相关滤波(KCF)跟踪算法在遮挡场景中出现跟踪性能降低甚至跟踪失败的问题,提出了一种核相关滤波和卡尔曼滤波(KF)预测相结合的模型自适应抗遮挡图像目标跟踪算法KCF-KF。首先,考虑到传统KCF目标跟踪算法中缺少遮挡评估的问题,通过引入响应图的峰值旁瓣比来对图像目标的遮挡情况进行判断,并将遮挡类型划分为部分遮挡和严重遮挡。其次,根据遮挡程度采取不同的模型更新策略,当目标无遮挡或者部分遮挡时,替代传统KCF跟踪算法中采用固定学习率更新模型的方法,通过自适应地调整模型学习率来更新目标外观模型,避免跟踪漂移;当目标被严重遮挡时,停止KCF模型更新。最后,应用严重遮挡之前的运动信息构建卡尔曼滤波器状态空间和位置输出模型,设计卡尔曼滤波算法预测运动目标轨迹来估计遮挡情景下的目标位置,从而解决在遮挡场景中目标跟踪失败的问题。采用OTB-2013标准数据集进行大量实验,结果表明:所提的混合跟踪算法KCF-KF的距离精度为0.796,重叠成功率为0.692。与其他传统跟踪算法相比,该混合算法的跟踪精度和跟踪成功率均优于其他算法,并且在遇到目标遮挡挑战时具有更好的跟踪性能,有效地解决了跟踪过程中的遮挡干扰问题。