Research On Three-dimensional Measurement Technology Of High Dynamic Range Surface Based On Polarization And Adjusting The Brightness Of Projection Pattern

Three dimensional(3D)measurement technology is widely used in archaeology,nondestructive testing,medical imaging and high-end equipment manufacturing With the development of science and technology,non-contact structured light 3D measurement technology has gradually become one of the main methods of 3D measurement technology due to its advantages of fast measurement speed,high precision and no need to contact the measured object.However,there are also some problems with structured light measurement technology.One of the most typical problems is the measurement of high dynamic range surface objects.Due to the large difference in reflectivity of the surface of the measured object,the image of the measured object captured by the camera will be too bright or too dark,resulting in measurement errors.When the surface reflectivity of the object is too high,the encoded pattern projected onto the measured object will be overexposed when captured by the camera.It is difficult for the camera to capture the correct coding pattern,resulting in errors in subsequent decoding,resulting in the loss of the final 3D point cloud.If the reflectivity of the object is low,the image captured by the camera will be darker,and the signal-to-noise ratio of such an image is too low,which will also cause a large measurement error and affect the accuracy of the 3D point cloud.To solve the above problem that it is difficult to perform 3D measurement of high dynamic range surface objects,this paper will introduce the camera response function technology based on the polarization system and the rapid adjustment of the brightness of the projection pattern,so as to eliminate the surface image of the measured object being too bright or too dark.Influence on 3D measurement,high dynamic range images can be obtained to improve measurement accuracy.And the proposed methods are all realized by quantitative analysis,which effectively guarantees the measurement efficiency.(1)Based on the principle of polarization extinction,a three-dimensional measurement method of camera intensity response function based on polarization is proposed.By exploring the imaging process of camera in polarization system,a model is proposed and defined:camera intensity response function based on polarization.By fitting the intensity response function of the camera under the action of the polarization system,the required polarization angle can be obtained.Finally,the best fringe image is obtained by combining the image fusion algorithm,so that the information hidden in the saturated region can be extracted to the greatest extent.Compared with the traditional polarization measurement method,this method can quantitatively calculate the required polarization angle,avoid the introduction of complex polarization bidirectional reflection distribution function model,and effectively improve the measurement accuracy and efficiency.And this method also solves the problem of too dark image caused by the traditional use of polarization filter to eliminate highlight.The experimental results show that the influence of the high signal-to-noise ratio between the black-and-white fringe and the edge of the measured object is significantly suppressed,and the experimental results show that the method can restore the high contrast between the black-and-white fringe.(2)Based on the principle of high dynamic range imaging by adjusting the brightness of projection pattern,a measurement method for quickly determining the best brightness of projection pattern is proposed.By projecting a uniform white light onto the measured object and capturing an image,the response relationship between the camera and the projector can be quickly calculated,and the pixel distribution curve of the response relationship can be drawn.Then,based on the proposed simple brightness selection strategy of the best projection pattern,the response relationship between the camera and the projector is quickly taken on the pixel distribution curve,and then the gray value of the best projection pattern is inversely solved.Finally,combined with the image fusion algorithm,the image with high signal-to-noise ratio and high contrast can be obtained to improve the measurement accuracy.Compared with the traditional method based on changing the brightness of projection patterns,the advantage of this method is that only one image is needed to quickly obtain the response relationship between camera and projector,so as to avoid the need to blindly project multiple patterns with different brightness.In addition,because this method proposes a strategy to determine the best projection pattern brightness,and combines this strategy with the image fusion algorithm,it avoids the calculation of homography matrix,greatly reduces the complexity of measurement and effectively improves the efficiency of measurement.(3)To verify the effectiveness of the proposed methods,structured light measurement systems are built for the two methods.In the experiment,the proposed method is evaluated by using many indexes such as signal-to-noise ratio,image contrast and frequency domain analysis.Finally,the measured object is reconstructed.The experimental results show that the proposed method can solve the problem of too bright and too dark surface image of the measured object,and effectively improve the three-dimensional measurement accuracy of high dynamic surface objects.And the above two methods are the best polarization angle or the best projection pattern brightness obtained through quantitative calculation,which greatly improves the measurement efficiency.