Bidirectional Markov Chain Monte Carlo Particle Filter for Articulated Human Motion Tracking

Tracking human motion from monocular video sequences has attracted a great deal of interests in recent years. The difficulty in solving this problem is largely due to the nonlinear property of human dynamics and the high dimensionality of the state vector space required to model human motion. Traditional particle filtering methods usually fail in this situation because the distributions they sample from are ill-defined. In this thesis we propose a novel tracking algorithm, namely the Differential Evolution - Markov Chain (DE-MC) particle filtering. It is based on the particle filter framework but makes substantial changes to its core, i.e. the sampling strategy. In this new approach, the Differential Evolution algorithm and the Markov Chain Monte Carlo algorithm are integrated, aiming at improving both the accuracy and efficiency in approximating the posterior distribution. Global optimization and importance sampling are spirits of the proposed method. To apply the DE-MC particle filter to articulated model-based human motion tracking, we also integrate multiple image cues including the area of silhouettes, color histograms and boundaries to measure the image likelihoods. We find the Fourier Descriptor (FD) to be a new and effective image feature in human motion tracking applications. Our other contributions, such as a modified color cue-based measurement function and a simple adaptive strategy for sampling, also help to improve the performance of the human tracker. Experimental results including the comparison with the performance of other particle filtering methods demonstrate the power of the proposed approach.

Download Full-text

Tracking Human Motion in Monocular Video Sequence with the De-Mc Particle Filter

10.32920/ryerson.14652045 ◽

2021 ◽

Author(s):

Du Ming

Keyword(s):

Markov Chain ◽

Differential Evolution ◽

Particle Filter ◽

Particle Filtering ◽

Motion Tracking ◽

Human Motion ◽

Image Feature ◽

Monte Carlo Algorithm ◽

Human Motion Tracking ◽

Monocular Video

Tracking human motion from monocular video sequences has attracted a great deal of interests in recent years. The difficulty in solving this problem is largely due to the nonlinear property of human dynamics and the high dimensionality of the state vector space required to model human motion. Traditional particle filtering methods usually fail in this situation because the distributions they sample from are ill-defined. In this thesis we propose a novel tracking algorithm, namely the Differential Evolution - Markov Chain (DE-MC) particle filtering. It is based on the particle filter framework but makes substantial changes to its core, i.e. the sampling strategy. In this new approach, the Differential Evolution algorithm and the Markov Chain Monte Carlo algorithm are integrated, aiming at improving both the accuracy and efficiency in approximating the posterior distribution. Global optimization and importance sampling are spirits of the proposed method. To apply the DE-MC particle filter to articulated model-based human motion tracking, we also integrate multiple image cues including the area of silhouettes, color histograms and boundaries to measure the image likelihoods. We find the Fourier Descriptor (FD) to be a new and effective image feature in human motion tracking applications. Our other contributions, such as a modified color cue-based measurement function and a simple adaptive strategy for sampling, also help to improve the performance of the human tracker. Experimental results including the comparison with the performance of other particle filtering methods demonstrate the power of the proposed approach.

Download Full-text

Genetic Operator-Based Particle Filter Combined with Markov Chain Monte Carlo for Data Assimilation in a Crop Growth Model

Agriculture ◽

10.3390/agriculture10120606 ◽

2020 ◽

Vol 10 (12) ◽

pp. 606

Author(s):

Alaa Jamal ◽

Raphael Linker

Keyword(s):

Monte Carlo ◽

Parameter Estimation ◽

Markov Chain ◽

Markov Chain Monte Carlo ◽

Data Assimilation ◽

Particle Filter ◽

Sequential Estimation ◽

Measurement Noise ◽

Genetic Operators ◽

True Value

Particle filter has received increasing attention in data assimilation for estimating model states and parameters in cases of non-linear and non-Gaussian dynamic processes. Various modifications of the original particle filter have been suggested in the literature, including integrating particle filter with Markov Chain Monte Carlo (PF-MCMC) and, later, using genetic algorithm evolutionary operators as part of the state updating process. In this work, a modified genetic-based PF-MCMC approach for estimating the states and parameters simultaneously and without assuming Gaussian distribution for priors is presented. The method was tested on two simulation examples on the basis of the crop model AquaCrop-OS. In the first example, the method was compared to a PF-MCMC method in which states and parameters are updated sequentially and genetic operators are used only for state adjustments. The influence of ensemble size, measurement noise, and mutation and crossover parameters were also investigated. Accurate and stable estimations of the model states were obtained in all cases. Parameter estimation was more challenging than state estimation and not all parameters converged to their true value, especially when the parameter value had little influence on the measured variables. Overall, the proposed method showed more accurate and consistent parameter estimation than the PF-MCMC with sequential estimation, which showed highly conservative behavior. The superiority of the proposed method was more pronounced when the ensemble included a large number of particles and the measurement noise was low.

Download Full-text