In order to solve the energy distribution of the skewed normal distribution problems in the actual star imaging process, and improve the positioning accuracy of star point center, a point spread function (PSF) correlation algorithm based on skewed normal distribution is presented. The proposed algorithm establishes the corresponding PSF according to the energy distribution of actual star point. It uses the relevant principles to find the PSF with the highest similarity with star energy distribution, and then by determining the corresponding PSF maximum position to realize the positioning of the central star. The experimental results show that under the simulating condition of star image Gaussian noise in N(0,0.001)and the star center random distribution within 1 pixel, the average positioning accuracy of skewed normal distribution PSF correlation method can reach 0.04 pixel, which is far less than 0.4 pixel from the centroid method and 1.03 pixel from the Gaussian surface fitting method. Experimental results show that the proposed algorithm is better than the centroid method and Gaussian surface fitting method, which has good anti-noise performance and stability, and improves the positioning accuracy of the star center.