| The high precision measurement method of the position and orientation of the remote object, is one of the hot points of vision inspection in the fields of aerospace, wheel alignment and so on. Monocular vision measurement is one of the measurement methods that obtain the physical dimension and position and orientation of object by single vision sensor. Monocular vision measurement has a number of advantages, such as simple in structure, less calibration steps. At the same time, this method also avoid the inadequate of stereovision, like small FOV(field of view), hard to match. On the basis of existent technologies, a brand new flat target is designed to measure the position and orientation of object. The new monocular vision measurement system is developed, which has many characteristics, such as simple structure, fast speed and good reliability. The main research work in this thesis includes:[1] Analyzing the existent monocular vision methods and their features, the value of this issue is expounded.[2] On the basis of the theory of vision inspection, a new monocular vision method is presented for getting the position and orientation of target;[3] In order to realize automatic camera calibration, a new feature-circle-based calibration drone is designed. And a set of algorithms for image processing, proved to be efficient, are proposed as well;[4] Under laboratory condition, infrared source is designed. The entire measurement software is programmed. And the monocular vision measurement system is built.The experiments of principle certification and measurement accuracy, are presented to prove the stability of this system. These experiments lay a solid foundation for wheel alignment system.[5] At the scene, the structure is adjusted to achieve satisfactory result. The experiments of repeatability are performed. The results show that, the repeatability precision of angles is less than 8″. The experiment data show that this measurement system is stable and reliable. |
PDF Full Text Request