Research And Implementation Of The 3D Reconstruction Based On Surface Structured Light Method

In recent years, the three-dimensional reconstruction technology,by which the 3D digital model of an actual object can be obtained, is widely used in the fields of industrial inspection, reverse engineering, biomedical, costume design and so on. The technology based on surface structured light is more extensively studied and applied because of its advantage of non-contact and high accuracy. When the 3D digital model of an object is reconstructed by the technology, the surface structured light is projected onto the object and modulated by the surface shape of the object. The image of modulated light is captured by a camera. After analysis of the image, the corresponding relationship is identified between the projection point and the imaging point about a point on the surface. According to the relationship, the depth of each point on the surface is calculated in order to reconstruct the 3D model of the object on the basis of the principle of triangulation.In the traditional 3D reconstruction systems composed of a projector and a camera, the constraints of the system configuration are very strict, so that it is difficult to place and adjust the position of the camera and the projector, and makes it complicated to calibrate the projector. These problems are researched and the following results are obtained:(1) Based on the analysis of the traditional 3D reconstruction system, a kind of phase-depth mapping method is proposed, which provides more freedom to place the projector and the camera and makes the configuration of the system more flexible;(2) In the study of the calibration of 3D reconstruction system, the algorithm proposed by Zhang Zhengyou is realized to calibrate the camera. The method for calibrating the projector is presented in which the calibrated camera is used to calibrate the projector without the special calibration equipment;(3) The algorithm is realized to extract the X-type corner points on the calibration board. And the method is researched to find out the central point of each color circle marker projected onto the checkerboard. For the points from the two methods mentioned above, an automatic method is proposed to sort these points according to the spatial relationship among them. By these means, the efficiency of the system calibration is largely improved.Finally, the phase is demodulated and unwrapped by the heterodyne phase-shift algorithm developed by our laboratory, and the three-dimensional surface reconstruction is completed according to the calibration results and the phase-depth mapping.