Image mosaicing by camera pose estimation based on extended Kalman filter

Official URL: http://dx.doi.org/10.1007/978-3-319-11758-4_49

We develop a sequential image mosaicing approach for aerial images of pseudo-planar scenes which is based on the estimation of camera pose from images. We use Extended Kalman Filter (EKF) to update the camera pose and scene parameters with every new image which improves the global consistency of the mosaic. Proposed approach is tested on aerial images where visually appealing results are obtained and residuals are quantified.