Used RTAB mapping, a powerful technique for creating accurate 3D maps using RGB-D cameras
Used the Kitti Dataset and explore how to use 3D lidars for object detection
Performed techniques for detecting objects in real-time, such as lidar-based segmentation and clustering
Used ROS2, to learn how to use rviz and PCL to create visualizations and analyze your point cloud data with ease. In addition, use explore cylindrical and planar segmentation, for extracting meaningful information from point cloud data.
Used LSD Slam using a monocular camera on a custom turtlebot with a camera mounted on top. Results were recorded in rosbag along with lidar and ultrasonic sensor data to be used for navigation. Virtual slam was further developed to be used on a larger robot like rover.
WhatsApp.Video.2024-06-26.at.00.mp4
The following video shows how the robot was moved in the real worls while generating the point cloud and performing SLAM.