The Airport Scene Surveillance Positioning System Based On Monocular Visual With Error Correction

In order to realize precise location of object in effective airport scene surveillance region,this text proposes the monocular vision coordinate measuring and positioning technology which is applied to the airport scene surveillance. First of all, the text starts from the homogeneous coordinates and the geometric transform concept. After that it establishes the pixel coordinate system, the image coordinate system, the camera coordinate system, the world coordinate system and the conversion relation matrix between them. After building the ideal mathematic model which is simplified by analyzing the internal and external parameters of the camera based on the characteristics in airport scene surveillance and the pinhole model,the objective measurement formula is obtainedHowever, measurement and errors always coexist. The process of error analysis refers to the random error and system error theory of the error theory. The study of the measurement uncertainty is based on the theories above. The dispersion which is characterized by the measurement uncertainty is used to analyze the image data obtained by preliminary experiments. It is proved that the positioning accuracy of the model meets the requirements in a specific area under the small field condition..In the condition of large field, the altitude error of the camera and the distortion induced by the camera lens are mainly considered. For the former, the 3D rotation matrix in the ideal mathematic theory model is replaced by a matrix added with angle of inclination ?angle of rotation ? to obtain a new measurement formula. For the later, by comparing the merits and drawbacks of the implicit-explicit method and grid method, the more suitable method,implicit-explicit method is applied to calibrate the distortion coefficients and get the relation model available.At last, combining the two correction methods, the experiment data which are obtained in large field environment are substituted in the model to verify its availability, correctness and reliability.