Authors:
Ivan Krešo
and
Siniša Šegvić
Affiliation:
University of Zagreb Faculty of Electrical Engineering and Computing, Croatia
Keyword(s):
Stereo Vision, Camera Motion Estimation, Visual Odometry, Feature Tracking, Camera Calibration, Camera Model Bias, Deformation Field.
Related
Ontology
Subjects/Areas/Topics:
Computer Vision, Visualization and Computer Graphics
;
Device Calibration, Characterization and Modeling
;
Image Formation and Preprocessing
;
Motion, Tracking and Stereo Vision
;
Stereo Vision and Structure from Motion
;
Tracking and Visual Navigation
Abstract:
We present a novel approach for improving the accuracy of the egomotion recovered from rectified stereoscopic video. The main idea of the proposed approach is to correct the camera calibration by exploiting the known groundtruth motion. The correction is described by a discrete deformation field over a rectangular superpixel lattice covering the whole image. The deformation field is recovered by optimizing the reprojection error of point feature correspondences in neighboring stereo frames under the groundtruth motion. We evaluate the proposed approach by performing leave one out evaluation experiments on a collection of KITTI sequences with common calibration parameters, by comparing the accuracy of stereoscopic visual odometry with original and corrected calibration parameters. The results suggest a clear and significant advantage of the proposed approach. Our best algorithm outperforms all other approaches based on two-frame correspondences on the KITTI odometry benchmark.