|
|
Hello,
for my thesis project, I need to develop a multi-agent SLAM package. I chose RTAB-Map as my reference package. I need to find a way to combine data from a drone fleet and send the updated data to each drone with a wireless network.
I want each drone able to perform SLAM autonomously. The ground station has only to improve the work.
Actually, my sensor suite consists of the Intel Realsense D435 depth camera and the T265 tracking camera.
I would like to ask if someone already did this, and, eventually, how.
For now, I tried something, but I'm not sure it's the right way to do it anymore.
I'll explain it.
I collected a ROS bag containing compressed depth images. I wrote a node that is able to start from compressed images and produce a point cloud. I want to do the same thing for every drone. Collect images, send them to a ground station, and convert each image into a point cloud. Then, I want to merge all the point clouds together and in some way send the final product to the drone to use it in real-time.
To take into account this PointCloud I thought to use the additional input of RTAB-Map.
Is this the best way to do it?
I tried, but I'm still not able to do it.
It is possible to make it work or I need another strategy?
Sorry if I wrote too much.
I hope it is possible to find an answer.
Thank you!
|
Locked
