We present an object tracking system based on an edge model forthe target characterization. The target position is estimated by lookingfor the model in the current image using a Hausdorff partial distance.Target is searched only in a sub-window of current edge image. Itsboundaries are determined by Kalman filter estimation that uses targetdynamics to predict the current position. We use a spiral searchingstrategy to find the actual position. The target model is updated ineach iteration by using unidirectional partial distance from the imageto the model. This model is refined by an enclosure operator in order toperform the target/background discrimination. The parameters of oursystem can be modified in an active way along the tracking task. Thesystem is shown to be robust to illumination changes and posevariations. The system has been also embedded in a mobile robot forpersonal robotics applications and integrated in a real-time OS (3 Hz)
展开▼
