| Robot technology is the current research focus of all countries. In the hot spots of robot research field, robot off-line programming is the key of utility of robot technology, and robot vision calibration is the premiss of off-line programming. The thesis, based on the study of PUMA560 robot basic theory and industrial application of national laboratory of industrial control technology, studies robot off-line programming and robot vision calibration. Based on these research achievements, made use of object-oriented idea and developed robot off-line programming and graphical simulation platform based on PC, furthermore, realized the robot vision-feedback control system. The primary achievements are following:1. Presented a novel camera self-calibration algorithm based on a plane pattern. In this algorithm a delicate method is firstly introduced to estimate the accurate center of projected circle with the use of equilateral-hexagon. Then a novel algorithm, based on the generally character of Euclidean space, is presented. This algorithm is not stringent to the camera's motion and is more flexible, for it doesn't need matching points between images. The contrast experiments indicate that the accuracy with the use of this method is comparable with the use of commonly method and the method needs less images. For its flexibility and facility, this method is very suitable to the fast calibration of robot vision.2. Presented a high-accurate camera calibration algorithm. The camera model in the algorithm is improved compared with Zhang's camera model. The model is also suitable when the tangential distortion is distinct in some lenses. Based on the model, an improved camera calibration algorithm is proposed. Through the proposed method, the initial parameters are solved by using the points near the image plane center. As the distortion is very litter near the image plane center, the solved initial parameters can be very close to exact ones, then, a subspace trust region method based on the interior-reflective Newton method is applied to solve all the camera parameters accurately. Both computer simulation and real datahave been used to test the proposed technique and very good results have been obtained. Compared with Zhang's method, the proposed method is more accurate and it worked better robustness.3. To meet with high accurate positioning of industrial robot, the traditional Two-Stages robot\eye calibration technique is improved based on the analysis of the principal calibration method of robot vision. The technique first makes use of two-stages method to solve the initial value of transition matrix, then establishes optimization mathematical model, Gauss-Newton method is used as its speediness in the optimization computation. The proposed technique not only possesses as many virtues as the traditional two-stages method, for example, automation, speediness, but also eliminates accumulative error brought on by two-stages method completely and improves the calibration accuracy greatly. Many experiments indicate that the proposed technique excels the traditional two-stages method distinctly.4. To solve the problems that the relation is not compact between the universal software and robot controller in initial off-line programming and graphical simulation system, we put forward and develop a new off-line programming and graphical simulation system for the robot, which based on object-oriented programming. The system has a friendly graphical interface. Off-line programming optimizes the programming environment and grounds for more intelligent robot programming. The object-oriented paradigm alleviates programming for robot task, model-based architecture make it easy for the system to be expanded.5. Aiming at recognizing and positioning ellipse, proposed an ellipse detection algorithm based on continuous edge features. The algorithm makes use of muti-methods to analysis image, and shows good robustness in complicated environment. Utilizing the continuity of edge features, The algorith...
|
PDF Full Text Request