Yongqi Li

Highly reflective surfaces often cause overexposure in fringe projection profilometry, leading to missing or distorted point cloud data. To overcome this limitation, this study proposes a mask-based method for generating adaptive fringe patterns, combining surface reflectance distribution characteristics with fringe projection principles. First, dynamic masks for non-saturated and saturated regions are generated according to pixel intensity values. Second, a linear regression equation is determined using the least squares method, and the light intensity scaling factor is calculated by combining the camera-projector coordinate mapping relationship to dynamically generate adaptive fringe sequences that match local reflectivity. Finally, the encoded fringes are projected onto the surface of the measured object for 3D reconstruction. Experimental results demonstrate that, compared with the traditional phase-shifting method, the point cloud coverage in highly reflective areas of typical metal workpieces increases significantly from 65.4% to 90.1%, with the use of only one set of masks, thereby improving measurement efficiency. Moreover, the proposed method exhibits good applicability for typical metal parts and plastics. The point cloud coverage in highly reflective regions shows an increase of over 22%.