The range image reconstruction algorithm of Geiger-mode APD laser radar system was studied, and a reconstruction algorithm based on pixel neighborhood kernel density estimation was designed. Starting from the system principle, the theoretical basis of the reconstruction algorithm of range image was studied with the detection probability model. According to the characteristics of the system, an improved algorithm based on pixel neighborhood kernel density estimation was proposed and its principle was analyzed. The histogram algorithm and the neighborhood kernel density estimation algorithm were verified by simulation data, and the range reconstruction accuracy rate curve was used for quantitative evaluation and comparison. The algorithm was further applied to real Geiger mode APD lidar data to reconstruct range image. The experimental results show that the reconstruction algorithm based on the statistical neighborhood kernel density estimation can effectively improve the reconstruction effect of the range image at low frame counts.