The target guarding problem (TGP) comprises two players, an evader and a pursuer. The evader tries to reach a stationary target while avoiding the pursuer, and the pursuer tries to intercept the evader before the target is reached. Optimal strategies for the TGP have been studied in detail. These strategies assume the availability of noise-free measurements of positions and speeds of the respective players. The pursuer may lose the game from a position of advantage du to lack of perfect data. In this work, a strategy is derived for the pursuer when the evader's position and speed measurements are corrupted by noise. A non-linear state space model is developed for the evader's maneuver. An Extended Kalman filter is then designed to estimate the evader's position and speed. This estimated data is used for calculating the probability of the target falling within the dominance regions of the pursuer or the evader. Based on this, a real time strategy is designed for the pursuer. Performance of this strategy is simulated and also validated through experiments conducted on a test-bed consisting of mobile robots. © 2019 European Control Association