The Research On Camera Calibration And3D Reconstruction Method In Visual Measurement

Computer vision technology has been widely applied in many fields, such asassocial production and life. It enables digital computer have the ability of recoveringthe surrounding environment and information from the two-dimensional data, and copyshuman visual cognition to computer. Camera calibration and3-D reconstructionmethods in visual measurement are key technologies of computer vision research.Thisarticle focuses on the two related theory.First of all, this article introduces the development and basic theory of the cameracalibration and3-D reconstruction in visual measurement, which includes commonfeature detection methods and camera imaging model, and analyzes the Zhang.Z.Y’scamera calibration theory.Then, study the realization principle of basic particle swarm optimizationalgorithm. For it is easy to fall into local convergence and the loss of global extremum,this paper improves the algorithm. Mutation operation is added in the progress ofoptimization, which improves the population diversity during the iteration process. Andthe related simulation are introduced, and show that the improved algorithm improvesthe global convergence and accuracy, and get better to adapt to the optimizationproblem of the multi-parameter and complex functions.Then, this paper transforms the calibration problem in the visual measurement intothe intelligent nonlinear calibration optimization problem. Based on Zhang.Z.Y’snonlinear imaging model, this thesis combines the particle swarm algorithm, anddesigns a new calibration method. Results of simulation and experiment show that,compared with the classic Zhang’s algorithm, the proposed method has high precisionand strong robustness. In addition, there is no limit to the number of calibration images,which provides great convenience for the staff of calibration work.Finally, this paper developed a3D reconstruction method based on binocular vision.And apply VC6.0and OpenCV programming tools to achieve the three-dimensionalreconstruction method. The experimental results verify the validity and feasibility ofthis reconstruction method.