This paper proposes a novel optical three-dimensional (3D) measurement method using the traditional space-time stereo system. In the proposed method, the projector not only shoots fringe pattern onto the measurement object to achieve precise matching, but also plays a vital role in the 3D information calculation. With the combination of two cameras and a projector, two digital fringe projection (DFP) measurement systems and one traditional space-time stereo measurement system can be obtained. In another word, the measurand will be measured three times simultaneously, which results in three independent point clouds of the same region of the object to be measured. So it is necessary to register these three sets of points for obtaining one final data set. The iterative closest points (ICP) method, which is known as the most popular registration approach, is sensitive to the initial estimation of the transformation between the two sets of points to be matched. Thus, a robust rough registration, which is introduced from Natasha, is useful for ICP to realize accurate registration. After registration, a scattered point set with redundant and errors, which are caused by overlapping, is obtained. Then some local surfaces are constructed for those overlapping regions using the moving least squares (MLS) method, and the points extracted from those surfaces are used to replace the points of the overlapping regions. Finally, a simplified, precise point cloud can be obtained.