The tracking performance of visual servo system is limited by the low sampling rate of camera. In this thesis, we present an adaptive prediction and interpolation method. Through this method, we could get the virtual sampling values of motion trajectory of target at high sampling rate. According to the high sampling rate trajectory, the servo motor could perform high speed tracking control. The results show that we could get more accurate motion trajectory by using high order linear model for adaptive estimation. In addition, the tracking performance of low sampling rate system in cooperated with the predictive interpolation will approach that of high sampling rate system.