Research And Application Of 3d Surface Measurement Technology Based On Multiple Frequency Phase-shift Technology Of Structured Light

The structured-light 3D shape measurement technology based on phase shift has been used in many kinds of fields, such as reverse engineering, product quality control, body measurement, heritage repair and so on. This technology grabs images with some phase-shifted grating to compute the wrapped phase which include 3D shape information. And then, using multiple frequency phase-shift technology or other coding method, the unwrapped phase used for 3D-reconstruction can be obtained.The structured-light measurement system has three parts: projection system, image acquisition system and software system. Based on triangulation, the system builds relationship between spatial coordination of object surface and unwrapped phase, and uses the unwrapped phase to get spatial coordination. Single and multiple frequency technology are the most used methods in phase shift technology. In this paper, based on simulation, we prove that compared to single-frequency technology, the multiple-frequency technology has an advantage in reducing random error and system error. Meanwhile we put forward the method of grating design and unwrapping process based on multiple frequency technology.The foremost unwrapping method based multiple frequency technology are look-up table and differential method, in this paper, these two methods are studied intensively. According to the study, unwrapping process of two methods is put forward. Based on computer simulation, we compare and analysis the unwrapping scene and allowance between two methods. Simulation result shows that look-up is better than differential method theoretically.Error compensation and calibration are key technologies of multiple frequency technology. Based on simulation of the impact of gamma nonlinearity of the projector on phase error, a method of phase error compensation for structural light measurement based on phase-shifting is put forward. In this method, binomial curve is used to approximate the intensity of the captured images, meanwhile, least square fitting method is used to compensate the phase error, and in this way, system error caused by the gamma nonlinearity of the projector is reduced greatly. Experimental results show that this method reduces greatly the phase error caused by the gamma nonlinearity of the projector. The average phase error is 5 times smaller than that before error compensation, meanwhile, the error rate of unwrapping based on look-up table method is reduced 14.5 times. This paper uses the method of growth of pixel segment to calibrate phase error. This method has advantages in space complexity and robust. After error compensating and calibrating, programming design of unwrapping is completed based on C++ program.At last, based on Theoretical analysis, experiments are performed to study the performance of Error compensation technology and error calibration technology, and to compare allowance, robust and unwrapping efficiency between look-up table and differential method. Meanwhile, some examples of the system application in projects are illustrated.