Target Motion Trajectory and Wall Surface Shape Estimation in Complex L-Shaped Corner Scene With Recessed Structures

Target detection and building scene perception in nonline-of-sight (NLOS) scenarios are critical challenges. This article explores the estimation of target motion trajectories and wall shapes, accounting for surface irregularities caused by structural elements like doors and windows, by leveraging the propagation characteristics of multipath echoes. Specifically, a noncoherent multipath imaging method is first employed for image-specific multipaths based on an analysis of the propagation characteristics and constraints at different motion moments. The resulting image is then segmented to identify distinct regions. The problem of target localization and wall reflection point estimation at any given moment is subsequently formulated as a Pareto optimization problem. By solving this problem, the target’s motion trajectory and the wall’s surface shape can be accurately estimated, even in the presence of surface irregularities. Finally, numerical simulations and experimental measurements are conducted to validate the effectiveness of the proposed algorithm.