Object trajectory prediction application to visual servoing

Visual Servoing is an important issue in robotic vision. Considering tracking as a particular case of visual servoing, motion estimation algorithms are used to predict the location of target and generate a feasible control input to keep the target in the center of the image. Several well known algorithms can be used for trajectory prediction such as Kalman filter, αβ/αβγ filters, circular prediction algorithms and so on, but in this paper, we present a new filter based on existing filters that improves the prediction made by any one of them. This new filter is based on parameter optimization of a fuzzy system, therefore, we have named it: Off-Line Optimized Fuzzy FILTER (OLOF FILTER). The robustness and feasibility of the proposed algorithm is validated by a great number of experiments and is compared with other robust methods.