Many real-world visual tracking applications have a high dimensionality, i.e., the system state is defined by a high number of variables. These kind of problems can be modeled as dynamic optimization problems, which involve dynamic variables whose values change in time. Most applied research on optimization methods has focused on static optimization problems and these methods often lack from explicit adaptive methodologies to tackle dynamic problems. Heuristics are specific methods for solving problems in the absence of an algorithm for formal proof. Metaheuristics are approximate optimization methods which have been applied to more general problems with significant success. On the other hand, particle filters lack from efficient search strategies. Particle filters are Monte Carlo methods which solve the sequential estimation problem approximating the theoretical distributions in the state-space by simulated random measures called particles. In this paper, we propose a general framework to properly hybridize heuristics/metaheuristics with particle filters. The aim of this framework is to naturally devise effective hybrid algorithms for visual tracking, guided by the use of abstraction techniques. Resulting algorithms exploit the benefits of both complementary approaches. As a particular example, a Memetic Algorithm Particle Filter (MAPF) is derived from the proposed hybridization framework. Finally, we show the performance of the MAPF when applied to a multiple object tracking problem.