Combine acoustics and vision into a "localization" module

dralley commented 9 years ago

This would be a useful change for several reasons.

It would allow us to keep a persistant model of our surroundings and our place within them, which is a technique used by Cornell, Sonia, and UF.
All of the sensor data from cameras, acoustics, (leddar, DVL, sonar?) can be combined into one unified model of the environment and reduce the amount of crap that mission control has to deal with.
It would make implementing new sensors much easier, since they can just contribute more information into that model instead of trying to handle it separately

smosh commented 9 years ago

Any info i can get to you in order to help with this?

dralley commented 9 years ago

It's in the very earliest stages of planning so I'm not quite sure. Since all the core acoustics stuff is going to be on another computer, most of it should be pretty independent of however we use the data after it gets sent back, so I don't think any of that really needs to change. The only thing that might be affected is the little bit of code that talks with the BeagleBone, which hasn't even been written yet I don't think. We can talk about it next meeting though.

Most of the work is going to be with mission and vision because right now mission is running vision code directly and it's very coupled together, and that's bad. Stripping that out and making vision independent (as part of "localization") is going to be fairly involved and nearly of the misisons will need to be changed to support it. Obviously this is going to happen in a separate branch :P

rafaelesteller commented 5 years ago

removed because too old

ncsurobotics / seawolf

Combine acoustics and vision into a "localization" module #26