Study Of 3D Point Cloud Acquisition Based On Laser Interference Fringe

Three-dimensional contour measurement technology plays an extremely important role in industrial manufacturing,military investigation,and medical imaging.In recent years,the optical 3D contour measurement has been a hot topic in the field of 3D contour measurement.The structured light projection method is a typical representation of optical three-dimensional contour measurement.It will project the grating stripe onto a surface of an object to be measured,and the camera will acquire the stripe image deformed by surfaces of an object,and the three-dimensional contour information will be obtained by demodulating phase information.Based on the three-dimensional contour measurement of structured light projection,laser interference fringes are used instead of general structured light to obtain the height information.In this thesis,because of the projection of laser interference fringes,optical filtering and lock-in imaging techniques can be used to eliminate ambient light interference.It has stronger capacity to resist ambient light interference,compared with the traditional structured light projection technology.On the other hand,due to the high brightness of the laser,the high resolution camera can be used to measure the threedimensional profile of large area objects.This is also not achieved by traditional grating projection measurement.In the process of obtaining the position data of the fringe image from the acquired fringe image,according to the feature that the fringes are parallel to each other,all the steps that are used in the image processing are one-line operations,and then extend to all the lines in the image.The thesis makes full use of the fringe feature to achieve the fringe position extraction,and effectively reduces the complexity of the image processing algorithm.A custom single-row band-pass filtering method is used in the process of image filtering,which makes it easier to extract the stripe skeleton in the filtered image.In the process of image calibration,multi-plane calibration method was used in this thesis.That is,the three-dimensional space to be measured is divided into a plurality of parallel planes according to the stripe pitch,and each plane is calibrated to obtain a corresponding calibration parameter,and finally it is expanded to become a calibration parameter of the entire three-dimensional space.Through experiments and corresponding analysis,it is found that there is a linear functional relationship between the calibration parameters of equally spaced parallel planes.The ideal reduction results are obtained,after fitting the calibration parameters linearly.In the process of three-dimensional contour reconstruction,the shortest diagonal stitching method is used to deal with the triangulation problem between adjacent two fringe curves,and then extend to three-dimensional space.The three-dimensional contour of the entire object to be measured is obtained,after drawing and displaying the object image by OpenGL.