Due to the low calibration accuracy for fisheye lenses by using circular control points, this paper proposes a fisheye camera calibration algorithm optimized by centroid method. Firstly, initial values of intrinsic and extrinsic parameters of imaging model were obtained, and then initial value of distortion parameters were solved by Two Step method. Secondly, centroid of ellipse detected from image and ellipse projected by imaging model were solved respectively. Then distance between those two centroid was minimized and cost function was established in order to optimizing the model′s parameters. Finally, parameters of imaging model were calibrated precisely. Experiment results with quantitive and qualitative comparison show that our method can improve the accuracy of calibration for fisheye lens.