In order to measure the Euler angle of the axis perpendicular to the ground level (yaw axis) of the triaxial air bearing table, a new approach for attitude estimation using an indoor star tracker is proposed. Considering the difference of the working condition between the indoor star tracker and the star tracker on orbit, this study is to improve the indoor-star tracker attitude estimation. The iterative attitude algorithm is used in the attitude estimation model of the indoor-star tracker to achieve a higher precision attitude. Because the iterative attitude algorithm requires a good guess of initial altitude matrix, an attitude optimization using the particle swarm optimization is introduced to get the initial attitude matrix. The simulation results show that the improved attitude estimation based on an indoor-star tracker can improve the attitude estimation precision of the yaw axis when measuring all three Euler angles efficiently.