Estimating the Kinematics and Structure of a Rigid Object from a Sequence of Monocular Images
IEEE Transactions on Pattern Analysis and Machine Intelligence
Recursive Estimation of Motion, Structure, and Focal Length
IEEE Transactions on Pattern Analysis and Machine Intelligence
Structure and Motion from Line Segments in Multiple Images
IEEE Transactions on Pattern Analysis and Machine Intelligence
| Hi-index | 0.00 |
The paper presents the new formulation for estimation of depth using Extended Kalman filter for the case of known/unknown motion parameters. The use of spherical projection, which is the hallmark of the present work, has simplified the formulations. Two approaches have been suggested: one in which, measurements are not part of the state vector, and another in which measurements are part of the state vector. When the measurements are included in the state vector, it leads to nonlinear state equation. On the other hand, when the measurements are not included in the state vector, it leads to nonlinear measurement equation. The second approach is also amenable to estimate more number of depths by increasing both the measurement and state vectors.The performance of the above approaches has been evaluated on a sequence of images of vase and cube. It is observed that results due to the approach where measurements are not part of the state vector are found to be converging and more close to approximately measured depth values. The Extended Kalman filter is more sensitive to initial values.