Shape acquisition using a color stripe pattern with stereo structure

Traditional coded structured light with spatial neighborhood is a well-known technique which is used to acquire 3D information. In this paper, we use stereo structure by adding a camera so that the calibration of projector could be omitted. By projecting a new color stripe pattern in which the thick stripes are encoded of De Bruijn sequences with six different channel values and order three, while the thin stripes that divide every thick stripe into two parts are encoded with multistep channel values, the images are grabbed with two cameras which are calibrated before. We use the Canny operator to detect the edge points which will be optimized later, then get the unique coded words of every edge point by using a twice decoding algorithm and correspond the edge points between two stereo images by these unique coded words from twice decoding. Finally, 3D information could be reconstructed using the principle of triangulation. In addition, we introduce three new pure color stripe patterns to solve the cross-talk problem which is caused by the DLP projector and CCD camera before the step of corresponding. The experimental results are quantitative and qualitative.