博主github:https://github.com/MichaelBeechan
博主CSDN:https://blog.csdn.net/u011344545
代码执行环境:Windows 8 + OpenCV3.0
https://github.com/MichaelBeechan/MyStereoLibviso2
Stereo visual odometry is a critical component for mobile robot navigation and safety. It estimates the ego-motion using stereo images frame by frame. In this paper, we demonstrate an approach of calculating visual odometry for indoor or outdoor robots(UGV) equipped with a stereo rig. Differ from others visual odometry, we use an improved stereo-tracking method that combines information from optical flow and stereo to estimate and control the current position of Unmanned Ground Vehicle. For feature matching, we employ the circle matching strategy in VISO-2. A high-accuracy navigation system is used to evaluate our results on challenging real-world video sequences. The experiment result indicates our approach is more accurate than other visual odometry method in accuracy and run-time.