Target enhanced and color preserved 3D reconstruction based on a multi-wavelength polarization line-structured light

Line-structured light 3D reconstruction is an active optical imaging technique that typically relies on highly coherent laser sources rather than conventional illumination. However, the monochromatic nature of laser light restricts the acquisition of complete surface characteristics. In this work, we present a line-structured light 3D reconstruction framework based on broadband composite white light. To address the measurement ambiguity arising from insufficient source intensity contrast, we introduce polarization imaging, which depends solely on the angle of incidence and surface properties, independent of source intensity. Furthermore, to mitigate the inconsistent behavior across different wavelengths under broadband illumination, we employ the Degree of Linear Polarization (DoLP) as the observation basis. We decouple the incident angle from the material response, and exploit the intrinsic relationship between wavelength and polarization to estimate the root-mean-square (RMS) surface roughness of the object. The estimated roughness cue is then incorporated to enhance target extraction during 3D reconstruction.