You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
step 3. Listen and save the pose x/y/z and roll/pitch/yaw information from rostopic '/stereo_odometer/pose'.
- question 1: From the video in test.bag, the road is flat and just have one 90-deg right turn, I know the position and attitude is relative to the camera frame but not the car body frame, but even so, I still think the position and attitude I got with big drift, as below graph of test 1.
step 4: repeat steps 2 and 3
- question 2: from the result of step 4, I find the position and attitude not only still with big drift, but also with big difference with test 1, as below graph of test 2, I am very confused.
- question 3: From the video in test.bag, I find the camera is face to the right side but not forward of vehicle, is it a good/helpful to improve the detection accuracy of algorithm?
- question 4: do you have some newer code or advices, need I change some parameters to solve question 1 and 2?
Hi, thanks for your viso2_ros, it's a great job!
Currently, I have some questions of viso2_ros and look forward your support, thanks in advance :-)
My environment: Ubuntu 18.04, ROS Melodic.
My steps and questions:
thanks a lot!