[ROS tutorial] RTAB-Map in ROS 101

RTAB-Map (Real-Time Appearance-Based Mapping) is a RGB-D SLAM approach based on a loop closure detector.
The loop closure detector uses a bag-of-words approach in order to determinate if a new image detected by an RGB-D sensor it is from a new location or from a location that it has been already visited.
Of course, this is a very summarized explanation, you will get more details on how this loop closure detector works inside this Course.
In this course you will learn how to perform RGB-D SLAM by using the rtabmap_ros package.
At the end of the course, the following topics will have been addressed:
- Basics of RTAB-Map
- How to use the rtabmap_ros package
- How does loop closure detection work internally
- How to generate a 3D Map of an environment
- How to perform Autonomous Navigation
Links and resources mentioned in this video:
Access the course: www.theconstruc...
#1: Intro to RTAB-Map (0:09)
#2: RTAB-Map ROS MicroCourse (8:34)
#3: Autonomous Navigation with RTAB-Map (24:59)
#4: Final Recommendations (48:06)
~-~~-~~~-~~-~
Please watch: "ROS Developers LIVE-Class Video Series" (Entirely practical online ROS Live-Class on How to Develop with ROS)
• ROS Developers LIVE-Cl...
~-~~-~~~-~~-~

[@walacerosalaurindo]

Great work!
I'm looking forward to watching ROS Perception in 5 days!

[@TheConstruct]

Thanks Walace! Working on it. Keep in touch

[@Uditsinghparihar]

Nice tutorial. But why the rtabmap_ros has not put such a precise tutorial in their documentation for beginners?

[@TheConstruct]

Ha ha ha... you should ask them!

[@sy2532]

can you update this example using Intel RealSense t265 and d435? This is a bit dated but still very good.

[@shrijank522]

Really great video for beginners. I was wondering if this RTAB map can be applied to an environment which has very fewer features. In such cases wheel odometry data becomes important. But is it possible to tune any parameters for such environment where features are significantly less in RTAB map having wheel odometry, kinect, 2d lidar?

[@muratkoc4693]

thanks for video. At navigation, Does the costmap constructed as LaserScan (original turtlebot3 costmap param) or PointCloud2?

[@achraf1647]

can i use it with Rplidar A1 ?

[@pramodkeshav7314]

How to do rtab_map with clearpath Husky robot with the realsense camera

[@malharjajoo7393]

How do you create a 3D Octomap from this ?

[@HasnainRaza-yr7fl]

Hi, I am really new to ROS and I am using ROS in my final year project. In my project, I am basically trying to traverse an unknown environment by creating a map of the environment and traversing the environment using the map both continuously; so not like creating the map first and then navigating using the map, but rather doing both the tasks together. Can you please guide me how should I do this? I am really stuck because all I can find on the web is how to do both the tasks independently and not together.

[@TheConstruct]

You need to create the map first. Then you need to localize the robot in that map. Finally you need to use path planning and obstacle avoidance to move the robot along the map. Watch this series of videos to understand how to do it: ruclips.net/video/mYwIu4OVMR8/видео.html
A better option is to take this course that teaches you step by step: www.theconstructsim.com/construct-learn-develop-robots-using-ros/robotigniteacademy_learnros/ros-courses-library/ros-courses-ros-navigation-in-5-days/

[@HasnainRaza-yr7fl]

Oh. So there is no method by which we can create an autonomous map making robot? By autonomy I am referring to using the already created part of the map to explore and map the unexplored territory

[@TheConstruct]

Yes, it is possible. What you mention is called 'exploration'. Check this video for an elegant implementation in ROS: ruclips.net/video/X-fmHZA6dHk/видео.html

[@keyserswift5077]

Can you make a tutorial on how to download it? I am using linux kubuntu but its a little new to me.

That is very helpful. Thank you very much.