Abstract: This paper first presents the principle of visual servoing,and then proposes a model independent uncalibration visual servo control method.Model independent uncalibration visual servo doesn’t need robot kinematics and camera models to control a robot.The control problem is formulated as a nonlinear least squares optimization,and the control law is deduced.The updating equation of image Jacobian matrix is also given.Finally,simulation results of tracking an object demonstrate the validity of the algorithms.