Research On Key Technologies Of Structured Light Binocular 3D Imaging

With the in-depth development of various research fields,three-dimensional measurement technology has been highly concerned by more and more researchers..In computer vision,medical image processing,archaeological exploration,modern industrial production and other aspects,three-dimensional measurement technology has its irreplaceable role,the application prospect is very wide.According to the obtained information,it is an important research content in the field of three-dimensional measurement to achieve high-precision three-dimensional reconstruction of the object surface.The 3D reconstruction method adopted in thesis is the binocular 3D reconstruction based on Gray code combined with phase-shifted structured light.In thesis,the relevant theories of the key technologies used in the process of 3D reconstruction are deeply analyzed,and related experiments are designed.First,the basic principles of camera imaging,monocular camera calibration and binocular joint calibration are analyzed,and the camera calibration experiment and error analysis are carried out using a black and white checker-board.For the binocular imaging system,the theory of binocular epipolar correction is also studied,and the correction of the left and right camera images is completed,which improves the speed and efficiency of matching corresponding points.Secondly,several common structured light coding methods are summarized and compared,and the three-dimensional reconstruction of the plaster David's head is carried out by using 5 Gray code patterns and 4 phase shift patterns.In the Gray code decoding process,a new decoding scheme was proposed to improve the problem of projected Gray code and anti-Gray code pattern decoding methods on the premise of correct decoding.Then,the content of 3D point cloud reconstruction and stitching is introduced in detail.For the reconstruction process,the method of stereo matching and constraint criteria,the elimination of measuring dead angles are introduced,and the principles of two three-dimensional coordinate calculation methods are analyzed.For point cloud splicing,the principle of the center extraction algorithm of concentric circle marks,the principle of automatic matching of mark points,and the calculation method of motion parameters between point clouds are introduced,and the effectiveness of the algorithm is verified by design experiments.After the three-dimensional point cloud is obtained,the motion parameters between the point clouds are calculated according to the identification points and the obtained point clouds are spliced,and good results are achieved.Finally,according to the selected hardware equipment,in the VS2010 environment,the C++language was used to independently de-velop a supporting 3D reconstruction software system based on MFC and OpenCV,and a good human-computer interaction interface was designed to make the 3D point cloud reconstruction process more concise and Efficient.The software system can also import point cloud file display in different formats,and can perform some simple operations,such as rotation,translation,etc.