Dear Authors,
I am currently working on implementing a relocalization mode for FAST-LIVO2. My initial idea is to first remove dynamic objects using ERASOR, then perform place recognition with iBTC, and finally estimate the robot's pose in the pre-built map by aligning the current odometry with the matched map location.
I would like to ask whether you think this is a reasonable solution, or if you have any suggestions for a more suitable relocalization pipeline for FAST-LIVO2.
Thank you for your time and advice.
Best regards,
PENG Yuhang
Dear Authors,
I am currently working on implementing a relocalization mode for FAST-LIVO2. My initial idea is to first remove dynamic objects using ERASOR, then perform place recognition with iBTC, and finally estimate the robot's pose in the pre-built map by aligning the current odometry with the matched map location.
I would like to ask whether you think this is a reasonable solution, or if you have any suggestions for a more suitable relocalization pipeline for FAST-LIVO2.
Thank you for your time and advice.
Best regards,
PENG Yuhang