Identification of Point Spread Function based on Turbulent Blur Model

To solve the complexity of blur in Aero-optics, we proposed a method of identification based on the turbulent blur model, namely an improved least-square point spread function identification method. Combined with the characteristics of the flow field, select the initial value in this method. Solve the second-order difference of the point spread function identified in the process of each two iterations, and observe its changes before and after. When the second-order difference tends to be stable, the iteration ends, and the iteration times at this time can be used as the iteration times of the least-square. The improved method can obtain the accurate result by improving the initial value and iteration times of conjugate gradient method. The proposed method is studied, and the experimental results illustrate the method's efficiency.