A new position and attitude measurement method based on unmanned aerial vehicle (UAV) image feature fusion and measurement system are proposed to precisely measure the position and attitude of UAV. To work out the position and attitude parameters of UAV, we take full advantage of the extracted robust corners, wing lines and symmetric axis of UAV. We compute the coordinates of corners by lines intersection, using least squares (LS) method. We find the unit direction vectors of wing lines and axis of symmetry by singular value decomposition method. Considering the corners wing lines and symmetric axis simultaneously, we get the position and attitude of UAV by the proposed feature fusion method. Numeric simulation results show that the precision of position less than 0.05 m and the precision of attitude less than 0.5°, when the measurement system contains two cameras, standard deviation of noise is 1 and the distance of UAV is approximately 1200 m. What′s more, increasing the number of network cameras helps to improve the measurement accuracy. Precision comparison proofs the correctness, feasibility and high precision of the proposed feature fusion method.