Hwasung Autonomous Driving Contest
2023.10.13~2023.10.14
I participated in the Localization Team, Vision Team, and the Path-Planning Team.
Localization Team
- Processing LiDAR Data
- 3D lidar points (pcd) → 2D laser scan: PCD to Laser Scan
- pcd(point cloud data) processing ex) clustering
- Detect parking space using lidar data
- algorithm
- clustering
- Setting up the IMU, GPS driver
Vision Team
- YOLOP: Lane Detection
- Post-processing algorithm
- TensorRT engine
- YOLOv5: traffic light, object detection
- Collecting dataset
- Lidar-Camera sensor fusion
- PersFormer: 3D Lane Detection
- CMake
- nms
- CUDA, torch, nvidia driver compatibility
- Lane detection algorithm
Path-Planning Team
- Morai Simulator
- Robot Localization package
- map, odom, base frame tf
- navigation stack
- writing roslaunch files
- HD Map
- QGIS: editing maps
- UTM, WGS84
- ROS Navigation Stack
- ROS service
- Goal manager logic