An adaptive window width centroid correction algorithm is proposed based on echo signal modeling for lidar distance detection. An adaptive window width model is established to obtain centroids based on the relationship between window width and saturation. The algorithm is adjusted by the median, and achieves high-accuracy time extraction of saturated waveforms. The simulation is carried out through Matlab, and the results show that when the SNR reaches 10 dB, the accuracy of the adaptive window width centroid correction algorithm is 0.3 ns. Compared with traditional waveform centroid algorithms, the accuracy of the proposed algorithm is improved by 92% and the problem of centroid drift can be effectively solved. The algorithm is verified utilizing board-level measured waveforms and is corrected for distortion of the measured waveforms, the results show that the time accuracy of algorithm can reach 0.5 ns , and ranging accuracy can achieve 7.5 cm in case of saturated waveforms. The proposed algorithm can effectively increase the dynamic range of ranging and reduce system complexity.