Missing information reconstruction method of single photon imaging lidar based on multi-echo extraction

The imaging results of Geiger-mode avalanche photodiode (GM-APD) laser detection and ranging (Lidar) is limited by the number of pixels and occlusion, and some details will be missing in the reconstruction results. A missing information reconstruction algorithm based on multi-echo extraction is proposed using Bump Hunting (BH) and spatial total variation features. Only a single GM-APD detector is used to achieve missing information reconstruction. The algorithm established a multi-echo extraction model by using the gradient feature distribution of echo, obtaining prior values of target position and waveform width, and then establishing a Gaussian fitting posterior model to assess prior accuracy and correct extraction results. For pixels with multi-echo, establish a local total variation feature optimization equation, complete the supplementation and spatial arrangement of multi-echo information, and achieve complete distance imaging. The resolution plate simulation results show that when N is 0.1–10, the proposed method can reconstruct the most missing information with good noise stability. The verification results of 900 m and 1 km long-distance outfield experiment data show that the proposed method can obtain twice the details of the Photon Processing for Active Single-photon Imaging (CASPI) and D Shin's methods. This study is significant for developing single-photon LiDAR target detection and recognition applications.