Three-Dimensional Profile Measurement System Based On Phase Measuring Profilometry

Phase measuring profilometry (PMP) is one of the most commonly used optical three-dimensional profile measurement techniques. It has been widely applied in industrial inspection, product quality control, reverse engineering, biomedical engineering, virtual reality, heritage conservation, human body measurement, etc. due to its advantages of fast-speed, automation, good-flexibility and whole-field measurement. In the last decade, PMP has been developing toward the direction of high-precision, high-stability and fast-speed measurement system. In this thesis, we mainly research the key algorithms in PMP to improve the precision and set up a three-dimensional profile measurement system. The main contents in the paper are as follows:(1) Phase calculation is one of the basic techniques of PMP. The point correspondences in the camera-projector system can be easily obtained by calculating the phase value from fringe patterns. We formulate the phase-shifting method and two phase unwrapping algorithms. After judging the pros and cons of different algorithms, we decide to choose multiple-frequency heterodyne as our phase calculation methods and use geometric constraints to improve the precision of point correspondence. Experimental results demonstrate that multiple-frequency heterodyne has a very good performance on phase unwrapping of complex objects which have discontinuous surface and complex reflectivity and texture. High precision can be reached through the point correspondence which uses geometric constraints.(2) In optical three-dimensional profile measurement system, the ambient light and non-sinusoidal fringe pattern caused by gamma nonlinearity of the projector are two main sources of phase error We analyze the camera image generation procedure and the cause of phase error. A new phase error model which considers the ambient light is proposed. Then we propose two phase error compensation methods for dark environment and ambient light condition respectively. Mathematic analysis and experiment results demonstrate that phase error changes in different ambient light conditions. The proposed methods can alleviate the phase error effectively.(3) System parameters calibration is one of the essential algorithms in optical three-dimensional profile measurement techniques. We introduce the camera parameters calibration method and analyze the pin-hole camera model and physical significance of its internal and external parameters. We treat the projector as a camera by establishing the DMD image of projector to utilize classic stereo vision calibration method to calibrate and optimize system parameters. Experimental results show that this method can achieve high precision of system parameters and three-dimensional measurement. The precision of our system is 0. lmm.(4) Generally, the camera and projector can be placed arbitrarily in the system based on stereovision principle, which means there is a fringe angle formed by the projector-camera baseline and the phase variation direction of fringe patterns. Consequently, the system sensitivity can be affected by the fringe angle. In this thesis, the relationship between fringe angle and system sensitivity is analyzed in both systems based on the triangulation and stereovision principles. Furthermore, the method of calculating the best fringe angle is proposed. We also indicate that a three-dimensional profile measurement system will be most sensitive to object depth change when the phase variation direction of the fringe patterns is parallel to the baseline, which is demonstrated by simulation and experiments.We develop a high-precision, high-resolution three-dimensional profile measurement system based on the camera-projector-camera structure. We use our system to reconstruct stationary objects, human hands and feet, human faces and eye corner areas. Reconstruction results demonstrate that our system has a very good performance and can be applied in different areas, such as industrial manufacture, clothing industry, plastic surgery, etc.