Research On Digital Fringe Projection Algorithm

Digital fringe projection three-dimensional topography measurement technology has become a representative technology in the field of optical three-dimensional measurement because of its non-contact,full field measurement,real-time,high-precision and many other advantages.It has a wide range of applications in industrial parts detection,ranging,reverse engineering,3D mapping,biomedical,target recognition,fashion design and other fields.This paper mainly focuses on gamma nonlinear phase error compensation and random phase-shifting phase extraction algorithm in digital fringe projection 3D measurement technologyIn the digital fringe projection 3D measurement technology,due to the nonlinear response between the projector and the camera,the projected grating fringe deviates from the standard sinusoidal characteristics.Therefore.there is a large phase error in fringe phase extraction,which is a main error source of 3D topography measurement.Cai et al.Proposed a phase compensation method based on the global phase error model,but this method needs to determine the model parameters through the system calibration in advance.Moreover,due to the multiple approximation processing used in the derivation of the error estimation,the phase error compensation is inconsistent with the actual situation,and there is a certain deviation.In view of the limitation of CAI method,based on the principle of minimum harmonic energy,a method to determine the optimal error model parameters by searching the extreme value of the evaluation function is proposed.The method compensates the phase error by determining the model parameters.Therefore,this method is a passive compensation algorithm,which does not need complicated pre-calibration processing.In addition,the calculation workload of extreme value search is greatly reduced due to the use of beetle antennae search algorithm.Simulation and experimental results show that this method can significantly improve the phase error compensation accuracy of CAI methodVargas applies principal component analysis to phase demodulation of interference fringes.Traditional PCA phase extraction algorithm is only suitable for equal step phase shift.At present,there are some limitations in the phase extraction algorithm based on PCA random phase-shifting.For example,only applicable to phase-shifting fringes with specific steps,the background light intensity can not be effectively eliminated,and the results of phase extraction are very sensitive to the nonuniformity of light field distribution.In view of the current research status,this paper proposes an ellipse fitting PCA phase demodulation algorithm with random phase shift of any number of steps The principal component extracted by PCA is orthogonalized by ellipse fitting.When the ellipse is modified,the ellipse centrifugal angle parameter is introduced to make the extracted phase accuracy more accurate.Considering the influence of uneven distribution of background light intensity and modulation coefficient on ellipse fitting,a normalized fringe ellipse fitting method is proposed.The influence of different processing modes on the accuracy of phase extraction is discussed,and the best mode scheme is proposed.The experimental results show that the proposed algorithm can effectively extract the phase of random phase-shifting fringes and deal with the uneven distribution of light field.The fringe processing algorithm originally used in interferometry is successfully extended to the digital fringe projection 3D topography measurement technologyThe experimental platform of three-dimensional topography measurement based on projection grating is built.The algorithm is developed and implemented by using MATLAB 2019b programming language.A large number of practical measurement experiments are carried out.The experimental and processing analysis results verify the effectiveness of this method.