A Novel Non-Line-of-Sight Imaging Enhancement Method of Time-of-Flight Camera

In recent years, Non-Line-of-Sight (NLoS) imaging methods based on Time-of-Flight (ToF) camera are mainly proposed for the point cloud data obtained by the double-scattering light path where the light source and the detector are distributed on opposite sides of the occluder. Meanwhile, the research objects are mostly static targets. In order to further improve the application capability of the ToF camera, the point cloud data of moving personnel acquired by the mainstream triple-scattering light path in the NLoS domain has attracted the attention of this paper. Moreover, a novel NLoS imaging enhancement method is proposed. By utilizing the depth distribution characteristics of point clouds and the principle of echo, the issues such as high noise and poor visual characteristics in NLoS point cloud data are addressed. The experimental results demonstrate the effectiveness of the proposed method. Compared to the Natural Image Quality Evaluator (NIQE) value of the raw data, the NIQE value of the enhanced results in this paper decreases nearly three times.