search

ROBOTIS-JAPAN-GIT / turtlebot3_slam_3d

Turtlebot3 3D-SLAM demo using RTAB-Map with Jetson TX2 and ZED Mini
170 stars 62 forks source link

Can Raspberry Pi3 replace Jetson TX2? #1

Closed dwkim31 closed 5 years ago

dwkim31 commented 5 years ago

I am using intel D435 and raspberry pie, remote pc.

I know Rtabmap works in the following way in Raspberry Pi3.

image

http://wiki.ros.org/rtabmap_ros/Tutorials/RemoteMapping

http://wiki.ros.org/rtabmap_ros/Tutorials/SetupOnYourRobot

But i think Raspberry Pi3 can't calculate yolo algorithms in this source files.

In this case, What parts of this source need to be modified so that the yolo algorithm can be computed on a Remote pc?

In other words, Is there any problem in compressing and importing the published image from intel d435 and running Detection image parts and Coordinates parts below figure?

image

Affonso-Gui commented 5 years ago

Hello,

There is no problem in running parts of the algorithm in a remote PC, which is easily enabled by the ROS framework.

Instead of launching turtlebot3_slam_3d.launch as a whole, launch each of its nodes separately. You can for instance launch turtlebot3_d435_bringup.launch and rtabmap.launch use_zed:=false rviz:=false from RasPi and darknet.launch use_zed:=false from your remote PC.

It is, however, advisable to have good networking environment and a CUDA-enabled PC.

If YOLO is still too heavy, you can change the model to yolov2-tiny.yaml in darknet_ros.launch or reduce the size in the yolov2.cfg.

By the way are you following this instructions to run D435 on RasPi? Since there is no USB 3.0 port the image quality seems to be quite low, which may affect the darknet detection accuracy.

Affonso-Gui commented 5 years ago

Great! Looking forward for more news.

If you have any more questions or problems please reopen or create other issues.