Research And Implementation On 3D Reconstruction About Oilfield Ground Equipment Of Uncalibrated Image

With the development of computer science and technology rapidly in recent years, 3d reconstruction has been widely applied in the cultural relic reconstruction, medical, defense and so on. But some problems such as 3d information acquisition difficult and fewer three dimension model exist in 3d reconstruction, so how to quickly and efficiently the reconstruct the 3d model has become the current hot research direction. This 3d reconstruction of not calibration digital images technology research is studied in this paper and applied in the reconstruction of oilfield ground equipment. Specific studies are listed as follows:First of all, the camera calibration is performed. With Zhang zheng you plane calibration method is used in the camera calibration. The camera's result was obtained, including camera parameter matrix and trajectory, and the precision of camera calibration result is validated.Then, the feature point extraction and matching is performed. Take pictures of oilfield ground equipment, harris corner detection algorithm was used to extract feature points and the NCC matching algorithm is used for feature points matching, a number of matching points points two-dimensional projection plane are derived.Secondly, the adaptive chaotic simulated annealing particle swarm optimization algorithm(ACPSO-SA) is proposed. The logistic sequence is applied for the initial population of chaos to speed up the convergence rate of the algorithm; The size of the inertia weight is adjusted according to the rate of change of population fitness to keep the global search ability and local search ability of the algorithm between the balance; Simulated annealing algorithm is introduced. As the inferior solution is accepted within limits temporarily in simulated annealing algorithm, so the algorithm jumps out of local optimum and the global optimal is achieved. At the same time the algorithm convergence analysis was conducted by using standard function. And The algorithm is compared with other intelligent algorithms.Again, the ACPSO-SA algorithm is used for fundamental matrix estimation. In order to solve the problem that the resistance of normalized eight point method is poor in the aspects of mismatching, first of all eight matching points are viewed as the minimal subset of fundamental matrix estimation and the normalized eight point method is used to estimate the corresponding fundamental matrix;Then the particle swarm algorithm is utilized to optimize a set of the fundamental matrix for getting rid of the error fundamental matrix caused bymismatching points and improving the accuracy of the algorithm, in this process, aiming at the shortcoming of particle swarm optimization(PSO) algorithm which is easy to fall into local minimum, the chaos characteristics,the adjustment mechanism of adaptive inertia weight and the simulated annealing algorithm are introduced into the particle swarm algorithm to improve the ability to the search of the algorithm in the paper. Practical application shows that the precision and computational efficiency of fundamental matrix estimation are improved.Finally, the 3d reconstruction of object is performed. In the previous research, match point by feature matching and fundamental matrix and camera parameters are combined to calculate three-dimensional coordinates of the space which is corresponded the two-dimensional point on the surface of the slides corresponding step by step. And 3d point cloud is obtained. Then the point cloud is used for Delaunay triangulation, three-dimensional space model is constructed and three-dimensional reconstruction is implemented...