aau-cns / mars_lib

MaRS: A Modular and Robust Sensor-Fusion Framework
Other
258 stars 34 forks source link

Asynchronous fusion of multi-sensor data #4

Closed WuZihao12 closed 1 year ago

WuZihao12 commented 1 year ago

Hi and thanks for your work! I want to study an asynchronous multi-camera system, is this code suitable for the purpose, is there any relevant case?

Chris-Bee commented 1 year ago

@WuZihao12 Thank you!

Yes, the MaRS framework is suitable for this case. All measurements can be processed asynchronously, MaRS will handle the intermediate IMU propagation steps and possible out of order measurement updates for you. In addition, with the modularity approach that MaRS is using, the handling of asynchronous measurment inputs is handled without the usual computational overhead. I hope this helps.

WuZihao12 commented 1 year ago

Thank you very much for your reply

WuZihao12 commented 1 year ago

Hi, can this module be used for optimization based vio? For example, vins-mono Or filter-based vio, such as open_vins.

WuZihao12 commented 1 year ago

Hello, is this visio a camera model?

Chris-Bee commented 1 year ago

@WuZihao12 MaRS would allow you to perform filter-based VIO with loosely-coupled vision. The vision sensor module that you found allows to incorporate loosely coupled vision with a vision world reference frame and scale of the pose as self-calibration states.

WuZihao12 commented 1 year ago

Thank you for your reply.

Chris-Bee commented 1 year ago

You are welcome.