Position and orientation estimation based on Kalman filtering of stereo images

The estimation problem of the position and orientation of a moving object from visual measurements is considered. Extended Kalman filtering of a sequence of stereo images is used to recursively compute an implicit solution to the projection equations. The proposed approach is general and can be applied to whatever number of cameras are fixed in the workspace. Computer simulations are presented to demonstrate the effectiveness of the algorithm in the presence of noise and to test the robustness of the estimate when some feature points are dynamically lost. Different types of geometric distortion as well as quantization and calibration errors are considered.