A restoration method for multi-frame stillness coherent lidar images is studied based on one-step predictive Kalman filter. Parameters of noise are estimated from the characteristic of the degradation images. In the image restoration progress of simulation, performance of the filter is studied under the condition of using less available images, as well as the limit value of the performance under the condition of using enough available images. The data are compared with those of an image averaging method. The results show that the method has a fast convergence speed, a higher improvement-in-signal-to-noise-ratio and a higher limit value.