Search

Visual object tracking remains a challenging task in computer vision although important progress has been made in the past decades. Particle filter (PF) is now a standard framework for solving non-linear/non-Gaussian problems, especially in visual object tracking. This study proposes an ant colony optimisation (ACO)-based iterative PF for object tracking. In the proposed method, the basic idea of ACO is used to simulate the behaviour of a particle moving toward the posterior distribution. Such idea is incorporated into the particle filtering framework in order to overcome the well-known particle impoverishment problem. An iterative unscented Kalman filter is used to design a proposal distribution for particle generation in order to generate better predicted sample states. For the likelihood model, the authors adopt the locality sensitive histogram to model the appearance of the target object, which can better handle the illumination variation during tracking. The experimental results demonstrate that the proposed tracker shows better performance than the other tracking methods.

Visual object tracking is a challenging research task in computer vision community which has been intensively studied by researchers in the past decades. Among all of the existing methods, particle filter based methods have gained special attention due to its ability to handle highly nonlinear/non-Gaussian multi-modality models. This paper proposes a robust particle filter based tracking method based on the Langevin Monte Carlo sampling. The Langevin Monte Carlo sampling method leverages the gradient of the posterior probability distribution to draw new particles. Meanwhile, an auxiliary momentum variable is introduced to ensure that the proposed sample cannot be trapped in local mode of the posterior distribution. As for the likelihood model, we introduce the locality sensitive histogram based model to handle the severe appearance variations induced by illumination change, partial occlusion or other factors. We compare the proposed method with several popular tracking methods from qualitative and quantitative perspectives. The experimental results show that the proposed method outperforms its counterparts.

Among many of the existing nonlinear filtering methods, particle filter has drawn special attention due to its ability to deal with nonlinear/non-Gaussian state estimation problems. But it suffers from degeneracy problem. In order the suppress this problem, in this paper, we introduced an iterated particle filter which is applied to the bearing only target tracking problem. The proposed method is on the base of the iterated Kalman filter which is used for building a proposal distribution for the particle filter. The iterated Kalman filter can improve the estimation accuracy by iteratively incorporating the current observations which makes it a better candidate for building proposal distribution and easy to be integrated into the particle filtering framework. Simulation results demonstrate that the iterated particle filter shows superior performance to the other methods in bearing only target tracking.

Sequential Monte Carlo method (also named as particle filter) is now a standard framework for solving nonlinear/non-Gaussian problems, especially in computer vision fields. This paper proposes an ant colony optimization (ACO) based iterative particle filter for visual tracking. In the proposed tracking method, the basic idea of ACO is used to simulate the behavior of particle moving toward the posterior density. Such idea is incorporated into the particle filtering framework in order to overcome the well-known problem of particle impoverishment. We design an iterative proposal distribution for particle generation in order to generate better predicted sample states. The experimental results demonstrate that the proposed tracker shows better performance than the other trackers.

Designing better proposal distributions can greatly affect the performance of the particle filters, which has been extensively studied in the literature. In this paper, we propose to design better proposal distribution using a new version of unscented Kalman filter- the iterated unscented Kalman filter (IUKF). The IUKF makes use of both statistical and analytical linearization techniques in different steps of the filtering process, which makes it a better candidate for designing proposal distribution in particle filter framework. Each particle is updated using an iterative manner. Through this process, the algorithm can make better use of the current observation for state estimation. To evaluate the performance of the proposed particle filter, we use a synthetic model and a real-world model for the experiments. The experimental results have shown that the proposed algorithm outperforms the alternatives.