查看文章

As conventional structured light with temporal and spatial encoded pattern is used to retrieve 3D shape, its measurement accuracy is inevitably affected by the broken and dislocated fringes or grids appearing in shaded, prominent or edge areas. In this study, a non-encoding fringe strategy is presented for large field 3D shape measurement under infrared light illumination. By utilizing the images recorded from several different orientations, the break and dislocation problems can be effectively solved by an algorithm of automatic identification and matching for out-of-plane height curves based on Euclid distance and quaternion. Meanwhile, the field of view can also be enlarged by prolonging the fringes in the edge areas. The examples of retrieving 3D shapes of the statues demonstrate the validation of the proposed approach.