Open shishiteng opened 3 years ago
Hi @shishiteng : Yes, especially in the case when VIO fails to accurately estimate the visual scale in the initialization process due to insufficient IMU excitation. Also, the depth data can be used to initialize the visual features' inverse depth and to add more constraints in the optimization process for state estimation. Some qualitative results are included in a paper under review, and I will release more results for detailed comparison. Stay tuned.
best,
@rising-turtle Got it! I am looking forward your new paper ~
Hi, thanks for sharing your code. I would like to know if rgbd-vio works better than vio, and how much performance rgbd-vio can improve?