Deep-learning Algorithm Collections for Autodrive with ROS
This project provides RGB image localization functionality based on AprilTag, suitable for scenarios where a cleaning robot needs to return to its charging station. Additionally, this project includes a stop function tailored for this scenario.
This project provides obstacle detection functionality based on the fusion of laser scanner and point cloud data, suitable for scenarios where a cleaning robot needs to navigate around obstacles.
Parking Slot Detection & Positioning Tool for ROS
vision-based lane & side tracking algorithm written in CPP with TensorRT using fscn model for ROS.
vision-based lane & side tracking algorithm (train code included) written in Python with PyTorch using fscn model.
Retrieve RTSP stream and publish to ROS node.
This project provides wall localization functionality based on laser scanners, suitable for scenarios where a cleaning robot performs wall-following cleaning tasks.