Fast-camera calibration of stereo vision system using BP neural networks
暂无分享,去创建一个
In position measurements by far-range photogrammetry, the scale between object and image has to be calibrated. It means to get the parameters of the perspective projection matrix. Because the image sensor of fast-camera is CMOS, there are many uncertain distortion factors. It is hard to describe the scale between object and image for the traditional calibration based on the mathematical model. In this paper, a new method for calibrating stereo vision systems with neural networks is described. A linear method is used for 3D position estimation and its error is corrected by neural networks. Compared with DLT (Direct Linear Transformation) and direct mapping by neural networks, the accuracy is improved. We have used this method in the drop point measurement of an object in high speed successfully.
[1] Yoram Yakimovsky,et al. A system for extracting three-dimensional measurements from a stereo pair of TV cameras , 1976 .
[2] Ken-ichi Funahashi,et al. On the approximate realization of continuous mappings by neural networks , 1989, Neural Networks.
[3] Geoffrey E. Hinton,et al. Learning internal representations by error propagation , 1986 .