Design and build an active stereo scanner and validation of results from a metrological viewpoint

3D model reconstruction by stereo vision is divided into three steps: the camera calibration, matching, and triangulation algorithm. The matching can be considered as the most important and effective part of accuracy. In passive stereo vision, matching is done by algorithms that are often based on image features. These algorithms yield poor results in low or high textured regions. The matching problem can be resolved by adding a light source to stereo rig. In this paper, a 3D scanner is designed and built by combining passive stereo and a slit laser. The match point is obtained as intersection between the epipolar line and the laser stripe on the image. It is solved by mathematical formulas on pixel scale. Then the accuracy and ability of the system are investigated by relevant standards and the measurement indexes. The test samples have been scanned over 180 degree and the accuracy, repeatability, reproducibility and reliability of the system are reported. The results show an acceptable accuracy for small industrial parts using the proposed system.