Neural Network-based Camera Calibration

Camera Calibration is the essential step of obtaining 3D information from views in the field of computer vision, which is widely used in the area of 3D reconstruction, navigation, visual supervision, etc. Therefore ,it is the base on the intelligent vehicle with vision functions. and the associative theoretic research has been put more and more attention in this field. This dissertation explores three kinds of technologies Camera Calibration ,Stereo Matching and Neural Network.Camera Calibration technology is a key step from 2D to 3D geometry information, and this technology plays an important role in Computer Vision . Camera Calibration ,that is to say , can get the inner parameters and outer parameters of the camera through experiment and calculation .Based on the analyses of the traditional Camera Calibration. the Camera Calibration method for binocular stereo vision is presented according to the characteristic of Neural Network and Camera Calibration ,In order to get the train data, the stereo matching has been done to the image and gray relative algorithm to realize the correspondence of point is used in the Stereo Matching。 The network is used to learn the relationships between the image information and the 3D information .and it neither requires the inner and outer parameters of the camera and any prior knowledge of the parameters .At the end of the dissertation, the full steps constructing a BP Neural Network are given , and predicting data and realistic data are analyzed and compared by using Matlab software in the experiment...