Structured light 3D measurement method is non-contact,low-cost and has been widely used in reverse engineering,face recognition,medical detection,industrial manufacturing and other fields.Among them,the phase-shifting structured light technology has received continuous attention because of its strong anti-interference ability,high precision,simple phase unwrapping process,and good real-time performance.Efficient and accurate encoding and phase unwrapping methods are key to improving measurement accuracy and enhancing system anti-interference ability,with the goal of reducing the impact of non ideal factors such as system noise and gamma response,and improving phase solving accuracy.The research work of this thesis is as follows:(1)For the calibration of the projector in the structured light 3D measurement system,the phase target method is used.Through the pixel coordinates and absolute phase values of the feature points in the camera imaging plane,the pixel coordinates of the corresponding homonymous points in the projector virtual imaging plane are calculated,which endows the projector with the ability to "shoot",and carries out sub-pixel optimization based on the cross ratio invariance.Finally,the projector is calibrated using the same method as the calibration camera.(2)An improved triple N-step phase shift reverse error compensation method is used to address the periodic phase error caused by the gamma response of the projector.On the basis of the N-step phase shift,two sets of sinusoidal grating fringes with different phase differences are projected.After reverse error compensation,the wrapped phase error obtained is significantly reduced.Applying this phase error compensation method to the projector calibration process also resulted in more accurate calibration results.(3)The phase unwrapping with and without reference planes was studied separately.If a reference plane is set in the three-dimensional measurement system and the depth range of the target to be measured is 2π within the phase domain,a phase free unwrapping method is used,which directly recovers the three-dimensional shape of the target to be measured in the phase domain from the wrapped phase difference between the target to be measured and the reference plane;If a reference plane is not set in the three-dimensional measurement system,an improved encoding and decoding method combining gray code and sine phase shift asymmetry is proposed,and a four-gray-level gray code is used to replace the two-gray-level gray code to reduce the number of projected encoding patterns.The periodic boundaries of the gray code and phase shift code are set at different positions and the farthest apart,which theoretically avoids the generation of periodic misalignment errors and improves measurement accuracy.(4)In order to verify the effectiveness of this method,a structured light 3D measurement system is built to calibrate the camera and projector respectively,and the calibration results that meet the accuracy requirements are obtained.Firstly,the phaseless unwrapping method was used for 3D measurement,and a good 3D morphology map was obtained,verifying the feasibility of this method.Secondly,the traditional gray code phaseshifting method and the proposed gray code asymmetric combination phase-shifting method with phase error compensation were used for actual measurement of planar and curved objects.The experiments showed that the encoding method proposed in this thesis has higher reconstruction accuracy,effectively suppresses phase errors,has good stability,and can highly restore the texture of the surface of the object to be tested. |