Research On 3D Surface Reconstruction Based On Structured Light

Three-dimensional reconstruction technology is an important research topic in the field of computer vision.Related technologies have been widely used in many fields such as industrial inspection,reverse engineering,cultural relic protection,virtual reality and 3D printing.How to accurately obtain the three-dimensional topography of the object surface and form the surface model of the object is a very meaningful task.Compared with the binocular vision system using passive feature matching,the structured light method is a three-dimensional measurement technique of active feature index,the projector projects a specific pattern of coding patterns to the surface of the object,then decodes and matches.And object which lacks the texture on the surface still can get accurate and robust matching result.Thus we can reconstruct the surface of texture-free object.This thesis researches and designs a three-dimensional reconstruction system based on time-encoded structured light technology.In this paper,firstly,we expound the principle of calibration of the system and how to implement it,and we establish the mathematical model of space point with camera and projector.Secondly,the principle of structured light encoding and decoding is introduced,and the implementation of the Gray code combined with the encoding inversion technology used in this system is described in detail.Thirdly,based on the two steps,the algorithm of 3D point cloud reconstruction is realized based on the principle of triangulation.The system obtains point cloud data with color and has high scanning accuracy.The integration of three-dimensional data is another aspect of 3D reconstruction.The initial point cloud data generated by the system is disorganized,structured surface formed by the three-dimensional discrete points can be applied in more scenarios.In this paper,an implicit surface fusion algorithm based on Poisson equation is implemented and the surface of the object is well reconstructed.