3D surface measurement plays an important role in robotics. In this paper, we provide a highly reflective surface measurement method based on combination of dual stereo monocular structured light system fusion. A predetermined rigid transformation matrix is utilized to improve fusion quality for the fusion area. Moreover, a final decision map is created to avoid the reconstruction of repeating regions of left and right system for ensuring the accuracy of the non-fusion area to achieve monocular system accuracy. The validity and accuracy of our algorithm are verified by designed experiments. Furthermore, various complex objects are measured including a metal surface, a phone case and a motor component, which show the proposed method has the capability to solve the highly reflective surface measurement problem and the occlusion problem.