Flatness is one of the most important properties for quality control of manufacturing mechanical parts. The most widely applied coordinate measuring machine techniques of measuring flatness error are ineffective to collect abundant sample points, and the probes must contact the tested surface. Existing noncontact optical techniques are not full-field measurement and their devices are complex. This paper presents a simple but noncontact full-field flatness measuring system based on fringe projection profilometry technique. The designed device projects fringe patterns onto the tested surface, and by calculating the phase of modulated fringe images and calibrating the phase-height mapping relationship, the full-field surface sample points are acquired. Polarizers are applied to eliminate intense highlight of the measured surfaces. Optimization algorithm is introduced to determine the reference ideal plane and flatness error is then calculated. Several experiments are conducted to demonstrate that the proposed flatness measuring system can be applied to general mechanical parts, and it has high precision and high repeatability.
In the current face 3D measurement technology, binocular stereo vision has been widely used. For the passive binocular 3D measurement system that does not need to project auxiliary light, it has the characteristics of simple system structure, but the result is not accurate enough and the algorithm is complex. Therefore, this paper proposes a fast measurement method for binocular stereo vision combined with infrared grating structure light. Because Digital-Light-Processing (DLP) projector has slow projection speed, dynamic images acquisition cannot be performed well, and when applied to the face 3D measurement, the eyes of the measured person will be stimulated by strong light, so a Micro-Electro-Mechanical System (MEMS) infrared projector is used in this paper. It has the advantages of high projection speed, high precision and no stimulation to the human eyes, so the MEMS projector can be well applied to 3D measurement of human face. The sinusoidal fringes are projected onto the face by the MEMS projector, and the phase is wrapped and unwrapped by phase measurement profilometry. In this paper, the four-step phase-shift method is used to calculate the wrapped phase, and the phase order is obtained according to the multi-frequency heterodyne principle. Fast matching of corresponding points of two image planes by combining epipolar and phase order constrained algorithms. The experiment verified that the highspeed, stable and low-cost face 3D measurement system was realized.