The hand-eye vision is used to measure the position and orientation of object, and provide the important information with the manipulation of the space robot, for example, capturing the target objects. Therefore it is significant to study the vision measurement method based on double hand-eye cameras.Firstly, the acquiring equipment and processing techniques of image are introduced and adopted in this thesis. Secondly, the camera calibration methods are studied, including linear calibration and nonlinear calibration. The method of nonlinear calibration is improved. Thirdly, the measurement method of pose, i.e. the reconstruction of the 3D object, is researched, involving point and line reconstructions. Finally, an idea is put forward to solve the measurement problem of the position and orientation under no guiding markers.A series of experiments are made in our laboratory and the department of automation at Anhui University, consisting image acquisition and processing, linear and nonlinear calibrations, and pose measurement. The experimental results show that the methods studied in the thesis are correct and effective.
|