With a footprint camera for surface imaging of laser spot and at the same time, it will cause the laser spot image in the camera to overlap with the ground image, resulting in poor center positioning accuracy of the laser spot. In this paper, according to the characteristics of the images measured by GF-7 satellite footprint camera, the ground noise image classification method is proposed to optimize the Gaussian fitting spot at the same time, the classification method of the spot image and the extraction accuracy of the corresponding spot center under the background of different ground noise are analyzed. Experimental results show that the proposed method has a center positioning accuracy of 0.11, 0.13, 0.16 pixel, and a positioning variance of 0.020, 0.262, 0.341 pixel, respectively, under different noise conditions of low, medium, and high. At night when the signal-to-noise ratio is better, the spot positioning accuracy is 0.02 pixel, and the positioning variance is 0.0036 pixel, indicating that the noise is the most important factor affecting the spot center positioning accuracy.