In order to realize real-time image detection in corneal topography, it is a key step to find an image sharpness evaluation function for fast, stable and accurate identification of real-time corneal images. For the optical imaging system of corneal topography, the defocusing effect on image sharpness is analyzed and a new image sharpness evaluation method is proposed. In the method, the Laplace function is selected as the coarse tuning evaluation parameter of a global image. Through Hough circle transform and sub-pixel edge detection, the center of the image is determined. With the center of the image, the clearest edge is found by the edge detection method based on the combination of Sobel operator and Canny operator. Based on that, the clearest corneal image is got in real-time. The experimental results verify that the proposed method has good stability and noise immunity, can evaluate the image sharpness accurately and effectively, and is suitable for real-time image detection in corneal topography.