| Structured light measurement is a non-contact three-dimensional measurement technology.By projecting a specific optical pattern(such as binary stripe or sine grating)on the surface of the object to be measured and capturing the image formed by reflected or scattered light with the camera,the three-dimensional shape information of the object surface is finally calculated according to the deformation or phase difference of the optical pattern.Structured light measurement has the advantages of non-contact,high precision and high efficiency.However,in the measurement process,the mutual interference between the projection light pattern texture and the reference plane,the camera imaging unit and the projector micro mirror array will lead to the appearance of moire patterns,which will affect the accuracy and reliability of the measurement.In order to solve the problem of moire interference in the projected light image in structured light measurement,an integrated calibration method based on collinear point target joint constraint is proposed in this paper.By simplifying the system calibration steps,the moire interference in the light plane constraint method is prevented; At the same time,a sinusoidal grating phase recovery method based on low rank decomposition was implemented,and the moire pattern of the sinusoidal image in the phase measurement method was separated through matrix decomposition.The main work content and research results are as follows:(1)A systematic integrated calibration method based on the joint constraint of collinear point target is proposed.A special collinear point target is used to construct a distortion measure function according to the straightness characteristics of projective geometry invariance,and optimize it to reduce the moire pattern of target image.At the same time,a bilinear interpolation algorithm is used to repair the center line of broken light stripe affected by moire pattern.Experiments show that when the standard deviation of image noise is less than 0.3 pixel,the absolute error of the distortion coefficient solved by the integrated calibration method proposed in this paper is kept within 0.1 pixel,and the re projection error is 0.1087 pixel.The whole process can be completed with only two projection target images,effectively reducing the molar interference in the linear structured light measurement system.(2)A moire pattern removal method combining a comprehensive frequency domain filter and a low rank matrix decomposition is proposed,and a comprehensive frequency domain filter is constructed to eliminate approximately periodic moire patterns.The low rank sparse matrix decomposition is used to separate the sinusoidal grating and moire interference components.The restored image is analyzed by peak signal to noise ratio and structural similarity index.Experiments show that,The method proposed in this article improves the peak signal-to-noise ratio of the restored image by 55.52 % compared to traditional image decomposition methods,with a structural similarity index of over 0.8.It can effectively eliminate sinusoidal grating moire patterns and reduce image distortion.(3)Analyzed the mixed sampling mathematical model of camera and projector,and obtained the selection principle of pixel size close to 1:1 for both.At the same time,equipment selection was carried out based on measurement accuracy requirements.Finally,good results were obtained by obtaining three-dimensional information on seamless steel pipes and plush toys,effectively proving the effectiveness and feasibility of the moire suppression method in this paper. |
