Abstract:<p indent=0mm>Backend trajectory optimization is an important part of the visual simultaneous localization and mapping system, which can significantly improve localization accuracy. However, the existing optimization methods based on the bundle adjustment have a large amount of calculation in large scenes and cannot be applied to end-to-end visual odometries. To solve this problem, a universal backend pose graph optimization algorithm with two visual odometries at the front end is proposed, which can be applied to end-to-end visual odometries. This method uses a high-speed but low-precision end-to-end visual odometry to run at high frequency, while a low-speed but high-precision visual odometry runs at a low frequency. Local optimization uses Gauss-Newton method iterative optimization through the constraints provided by two odometries. Global optimization is performed simultaneously which based on key frames scene matching. Experiments show that the simultaneous localization and mapping system which apply this optimization method can run in real-time on the KITTI dataset. Compared with the two visual odometries, the accuracy has been greatly improved. And compared with several well-known open source simultaneous localization and mapping methods that apply backend trajectory optimization, low errors have been achieved in trajectory error, absolute translational error, rotation error and relative pose error, taking into account the advantages of the accuracy of traditional methods and the advantages of high speed end-to-end methods. In addition, the optimization framework can also be applied to other more visual odometries.

Bi-objective Optimization for Robust RGB-D Visual Odometry

Globally Optimal Consensus Maximization for Robust Visual Inertial Localization in Point and Line Map.

Deterministic Optimality for Robust Vehicle Localization Using Visual Measurements

Robust visual odometry based on front and back cameras

High-Performance Visual Odometry with Two-Stage Local Binocular Ba and Gpu

Feature Regions Segmentation Based RGB-D Visual Odometry in Dynamic Environment

Robust stereo visual odometry using weighted and sparse optimization

6D Pose Estimation for Textureless Objects on RGB Frames using Multi-View Optimization

RGB-D Inertial Odometry for Indoor Robot via Keyframe-based Nonlinear Optimization

Robust Real-time RGB-D Visual Odometry in Dynamic Environments via Rigid Motion Model

Image Gradient-based Joint Direct Visual Odometry for Stereo Camera

A SLAM Pose Graph Optimization Method Using Dual Visual Odometry

Robust Stereo Visual Odometry Using Improved RANSAC-Based Methods for Mobile Robot Localization

Robust RGB-D Odometry Using Point and Line Features

Robust Stereo Visual Odometry Based on Probabilistic Decoupling Ego-Motion Estimation and 3D SSC

Robust and Efficient Visual-Inertial Odometry with Multi-plane Priors.

Stereo-based Multi-motion Visual Odometry for Mobile Robots

A Robust and Integrated Visual Odometry Framework Exploiting the Optical Flow and Feature Point Method

Stereo Orientation Prior for UAV Robust and Accurate Visual Odometry

Application Of Nonlinear Optimization Method In Stereo Visual Odometry