Figure - available from: Multimedia Tools and Applications
This content is subject to copyright. Terms and conditions apply.
Source publication
Generally, particle filters need a large number of particles to approximate the posterior for the purpose of ideal effect. Previous methods extract remarkable particles from the particles at time t-1 by nonlinear function. Those methods use the remarkable particles to reduce the number of particles and improve the accuracy of particle filter. Howev...
Similar publications
Recent research has provided several new methods for avoiding degeneracy in particle filters. These methods implement Bayes' rule using a continuous transition between prior and posterior. The feedback particle filter (FPF) is one of them. The FPF uses feedback gains to adjust each particle according to the measurement, which is in contrast to conv...
Most of the correlation filter based tracking algorithms can achieve good performance and maintain fast computational speed. However, in some complicated tracking scenes, there is a fatal defect that causes the object to be located inaccurately, which is the trackers excessively dependent on the maximum response value to determine the object locati...
Citations
... An error ellipse is constructed in accordance with the error covariance matrix of particles during resampling, based on which the particles are screened and optimized. A PF algorithm with significant local resampling is put forward in the literature [22], which uses a weight threshold and a distance threshold to extract significant local particles for MTT and to improve the accuracy of PF in the case of small particles. At the same time, the particle diversity is improved by adding Gaussian noise for particle roughening. ...
In this paper, a moving target tracking (MTT) algorithm based on the improved resampling particle filter (IRPF) was put forward for the reduced accuracy of particle filter (PF) due to the lack of particle diversity resulting from traditional resampling methods. In this algorithm, the influences of the likelihood probability distribution of particles on the PF accuracy were firstly analyzed to stratify the adaptive regions of particles, and a particle diversity measurement index based on stratification was proposed. After that, a threshold was set for the particle diversity after resampling. If the particle diversity failed to reach the set threshold, all new particles would be subjected to a Gaussian random walk in a preset variance matrix to improve the particle diversity. Finally, the performance of related algorithms was tested in both simulation environment and actual indoor ultrawideband (UWB) nonline-of-sight (NLOS) environment. The experimental results revealed that the nonlinear target state estimation accuracy was maximally and minimally improved by 12.83% and 1.97%, respectively, in the simulation environment, and the root mean square error (RMSE) of MTT was reduced from 17.131 cm to 10.471 cm in actual UWB NLOS environment, indicating that the IRPF algorithm can enhance the target estimation accuracy and state tracking capability, manifesting better filter performance.
... However, due to the difficulty of the occlusion problem, there are few research reviews on the anti-occlusion target model. Literature (Zhao et al., 2017) elaborated on the basic ideas of various tracking algorithms in occlusion scenes and the occlusion processing function of the target model, but the research object is the early tracking algorithm. In the literature (Wu et al., 2013;Sliti et al., 2018), when describing mainstream improvement schemes for the insufficiency of various tracking algorithms, there are few discussions on anti-occlusion schemes, and the essential function analysis of anti-occlusion is lacking. ...
Aiming at the existing problems that object tracking algorithm fails to track under the influence of occlusion conditions, the paper has improved the Kernel Correlation Filter algorithm. Firstly, the occlusion condition has been added to the Kernel Correlation Filter algorithm. If there is no occlusion, the Kernel Correlation Filter algorithm has used for object tracking. If there is occlusion, the improved algorithm based on Unscented Rauch--Tung--Striebel Smoother has been used. Secondly, the predicted position of the object has been feedback to the Kernel Correlation Filter algorithm. Finally, the combination of adaptive multi-model has been realized by combining the color histogram with the Kernel Correlation Filter algorithm, and the sparse representation method has been introduced into the training process to heighten the stability of the proposed object tracking algorithm. The experimental results using the proposed method on the OTB-2013 dataset can express that the proposed object tracking algorithm can reduce the occlusion interference in the object tracking process, and ameliorate the accuracy rate and success rate.
... With the development of target tracking, many visual target tracking algorithms have been investigated [13]. e mainstream target tracking algorithms include traditional tracking methods, correlation filtering-based tracking methods, and tracking methods based on deep learning [14,15]. ...
The structured output tracking algorithm is a visual target tracking algorithm with excellent comprehensive performance in recent years. However, the algorithm classifier will produce error information and result in target loss or tracking failure when the target is occluded or the scale changes in the process of tracking. In this work, a real-time structured output tracker with scale adaption is proposed: (1) the target position prediction is added in the process of target tracking to improve the real-time tracking performance; (2) the adaptive scheme of target scale discrimination is proposed in the structured support to improve the overall tracking accuracy; and (3) the Kalman filter is used to solve the occlusion problem of continuous tracking. Extensive evaluations on the OTB-2015 benchmark dataset with 100 sequences have shown that the proposed tracking algorithm can run at a highly efficient speed of 84 fps and perform favorably against other tracking algorithms.
... At present, the object tracking algorithm can be classified into two categories: discriminant model and generative model. The generative model had been mainly used it to describe the characteristics of object, and minimized the reconstruction error by searching for candidate objects, such as Particle Filter Tracking algorithm (Zhao et al. 2017), Mean Shift algorithm (Hsia et al. 2016), Kalman Filter (Vojir et al. 2014) and Mean Shift Fusion method (Jeong et al. 2017). However, this method often ignores the background information, so it is easy to produce drift phenomenon when the object had changed drastically * Jin Wang jinwang@csust.edu.cn ...
In order to enhance the robustness to complicated changes of multiple objects and complex background scene, the visual object tracking algorithm based on Adaptive Combination Kernel has been proposed in the paper. The object tracking procedure has been decomposed into two subtasks: Translation Filter and Scale Filter to estimate the object’s details. Firstly, the Translation Kernel Tracker has used the adaptive combination of Linear Kernel Filter and Gaussian Kernel Filter. The objective function has been developed to obtain the weight coefficients for Linear Kernel filter and the Gaussian Kernel filter, which incorporates not only empirical risk but also maximum value of response output for each kernel. The Adaptive Combination Kernel has the advantages of both local kernel and global kernel. Secondly, the tracking position has been calculated according to the response output of adaptive combination kernel correlation filter. Thirdly, according to the maximum response value, the scene-adaptive learning rate has been designed in the translation filter. The translation filter can be updated with the adaptive learning rate. Finally, one-dimensional scale filter has been used to estimate the object scale. The extensive experimental results have shown that the proposed algorithm is optimal on OTB-50 dataset in success rate and distance precision parameters, which is 6.8 percentage points and 4.1% points than those of KCF and is 2.0 percentage points and 3.2% points than those of BSET. The proposed algorithm has better robustness to the deformation and occlusion than others.
... Now, according to (2), the state estimation problem can be iteratively solved via prediction and update steps. In resampling step, the number of low-weighted particles is decreased and the number of particles with high weights is increased [47]. ...
Person re-identification, having attracted much attention in the multimedia community, is still challenged by the accuracy and the robustness, as the images for the verification contain such variations as light, pose, noise and ambiguity etc. Such practical challenges require relatively robust and accurate feature learning technologies. We introduced a novel deep neural network with PF-BP(Particle Filter-Back Propagation) to achieve relatively global and robust performances of person re-identification. The local optima in the deep networks themselves are still the main difficulty in the learning, in despite of several advanced approaches. A novel neural network learning, or PF-BP, was first proposed to solve the local optima problem in the non-convex objective function of the deep networks. When considering final deep network to learn using BP, the overall neural network with the particle filter will behave as the PF-BP neural network. Also, a max-min value searching was proposed by considering two assumptions about shapes of the non-convex objective function to learn on. Finally, a salience learning based on the deep neural network with PF-BP was proposed to achieve an advanced person re-identification. We test our neural network learning with particle filter aimed to the non-convex optimization problem, and then evaluate the performances of the proposed system in a person re-identification scenario. Experimental results demonstrate that the corresponding performances of the proposed deep network have promising discriminative capability in comparison with other ones. © 2018, Springer Science+Business Media, LLC, part of Springer Nature.
... However, it remains a challenging task due to many factors such as fast motion, in-plane rotation, out-of-plane rotation, shape deformation and occlusion. Many methods have been proposed and achieved excellent performance, which can be categorized into generative [1][2][3] and discriminative approaches [4][5][6][7][8][9][10][11][12][13][14][15][16]. The generative approaches include the particle filtering [1], sparse coding [2] and mean-shift [3] tracking methods. ...
... Many methods have been proposed and achieved excellent performance, which can be categorized into generative [1][2][3] and discriminative approaches [4][5][6][7][8][9][10][11][12][13][14][15][16]. The generative approaches include the particle filtering [1], sparse coding [2] and mean-shift [3] tracking methods. ...
... Zhao et al. [1] proposed a particles filter tracker based on a new resampling scheme which used a weight threshold and a distance threshold to extract remarkable local particles. Based on the particle filter framework, Zhang et al. [2] formulated object tracking as a structured multitask sparse learning problem which performs well in occlusion, drastic illumination changes and large pose variations. ...
Robust object tracking is a challenging task in multimedia understanding and computer vision. The traditional tracking algorithms only use the forward tracking information while neglecting the inverse information. An inverse relocation strategy is used to learn the translation and scale filters in the proposed tracking algorithm. To begin with, we learn a translation filter using both the forward and the inverse tracking information based on the ridge regression. The object position can be attained using the translation filter by the inverse relocation strategy. Secondly, the scale filter can be attained using the ridge regression and a smooth strategy is adopted to integrate the forward and inverse scale factors. Experiments are performed on the scale variation dataset and the OTB-50 dataset. Extensive experimental results show that the proposed algorithm performs favorably against several state-of-the-art methods in terms of precision and success rate. Meanwhile, the proposed algorithm is also robustness to the deformation to a great extent.
... In recent years, the sequential Monte Carlo methods, more known as particle filter, are considered as the algorithms of choice for visual tracking, it is a Bayesian sequential importance sampling technique used for posterior distribution estimation of state variables in a dynamic system under nonlinear, non-Gaussian environments [3,46]. The key idea is to represent the required posterior density function by a set of random samples (particles) with associated weights and to compute the estimates based on these samples and weights. ...
In visual tracking topic, developing a robust tracking method is very challenging, seen that there are many issues to look at, particularly, fast motion, target appearance changing, background clutter and camera motion. To override these problems, we present a new object tracking method with the fusion of interacting multiple models (IMM) and the particle filter (PF). First, the IMM is applied with a bank of parallel H∞ filter to estimate the global motion, the target motion is efficiently represented using only two parametric single models, and an adaptive strategy is preformed to adjust automatically the parameters of the two sub models at each recursive time step. Second, the particle filter is performed to estimate the local motion, we fuse the color and texture features to describe the appearance of the tracked object, we use the alpha Gaussian mixture model (α-GMM) to model the color feature distribution, the parameter α allows the probability function to possesses a flatter distribution, and the texture feature is represented by the distinctive uniform local binary pattern histogram (DULBP) based on the uniform local binary pattern (ULBP) operator; we fuse then the two features to represent the target’s appearance under the particle filter framework. We conduct quantitative and qualitative experiments on a variety of challenging public sequences; the results show that our method performs robustly and demonstrates strong accuracy.
Recently, sparse representation has been applied in object tracking successfully. However, the existing sparse representation captures either the holistic features of the target or the local features of the target. In this paper, we propose a dual-scale structural local sparse appearance (DSLSA) model based on overlapped patches, which can capture the quasi-holistic features and the local features of the target simultaneously. This paper first proposes two-scales structural local sparse appearance models based on overlapped patches. The larger-scale model is used to capture the structural quasi-holistic feature of the target, and the smaller-scale model is used to capture the structural local features of the target. Then, we propose a new mechanism to associate these two scale models as a new dual-scale appearance model. Both qualitative and quantitative analyses on challenging benchmark image sequences indicate that the tracker with our DSLSA model performs favorably against several state-of-the-art trackers.
