3D Reconstruction Of The Coded Structured Light Based On The Gray-Code And The Phase-Shift

3D reconstruction technique is a tributary research field of computer vision, which interlaces with computer graphics in implementation. This technique is witnessing promising prospects in such realms as production automation, robot vision, cultural heritage preservation, virtual reality and computer-aided medical science etc.. It is a tough task to accurately extract surface information of an object in a swift manner and to reconstruct the observed object using such information.This paper puts much emphasis on the study and design of a binocular stereo vision measuring system based on the Gray Code and Phase Shifting encoding method. A thorough analysis has been made on the 3D reconstruction theories and methods on coded structured light. After that, principles and the implementation of the joint calibration of double cameras are described, to further facilitate the construction of the geometrical constraints between the camera views and projective patterns. Meanwhile the algebraic description is modeled. In addition, various pattern coding methods are classified and compared, and a coding method is designed based on 7-bit Gray Code and 16-time Phase Shifting. The image capture, the edge extraction, the decoding and the computation as well as the reconstruction of the 3D point clouds come afterwards with software and hardware implementations. Consequently, the feasibility test in lab environment is made for the given system and the 3D point clouds of the object are successfully reconstructed. The merits of the system can be categorized as simple in coding, easy to implement, time-efficient and highly accurate.