Video tracking based on particle filters: Particle allocation, graphical models & performance evaluation

We develop novel solutions for particle filtering on graphs. An exact solution of particle filtering for conditional density propagation on directed cycle-free graphs is performed by a sequential updating scheme in a predetermined order. We also provide an approximate solution for particle filtering on general graphs by splitting the graphs with cycles into multiple directed cycle-free subgraphs. We utilize the proposed solutions for the applications of visual tracking using high-order Markov chain and distributed multiple object tracking.;We present a novel particle allocation approach to particle filtering which minimizes the total tracking distortion for a fixed number of particles over a video sequence. We define the tracking distortion as the variance of the error between the true state and estimated state and use rate-distortion theory to determine the optimal particle number and memory size allocation under fixed particle number and memory constraints, respectively. We subsequently provide an algorithm for simultaneous adjustment of the proposal variance and particle number for optimal particle allocation in video tracking systems. We further extend our work to multiple object tracking and articulated object tracking to allocate particles among multiple objects (object parts) as well as multiple frames.;We present a novel consistency measurement called multifold consistency to assess the performance of a chosen tracking algorithm computed recursively over a multifold of forward and backward frames. Our hypothesis is that the object states before and after this process should be consistent for a successful tracking and a disagreement in states indicates a possible error. Given ground truth, multifold consistency is used to evaluate the performance of different tracking algorithms as well as illustrate at which parts of the video one tracking algorithm performs well or poor. Moreover, we examine the robustness of tracking algorithms against possible errors by using perturbation analysis. Furthermore, we present an adaptive object tracking algorithm utilizing this score of each separate object to adjust the complexity of the corresponding core trackers, such as particle filters and mean-shift variants, and switch between these methods recurrently to extract the most reliable object tracks.