Title

• 국문: 복셀 3차원 좌표 기반 실시간 맵핑 및 위치추적의 SLAM 기술개발 

• English: Development of SLAM Technology for Real-Time Mapping and Localization Based on 3D Voxel Coordinates 

Funded by: 중소기업청(중소벤처기업부)

Period: 2024-08-01 ~ 2025-07-31

Role: participant

Manager:

Project summary:

Voxel-Based 3D Coordinate Mapping and Global Map Generation
Convert LiDAR point cloud data into voxel-based 3D coordinate structures.
Represent spatial attributes such as shape, size, orientation, and position in a volumetric format.
Implement adaptive voxel density control and apply Marching Cubes algorithms for surface compression and data simplification.

Development of LiDAR-Based Embedded System
Build an embedded control and computation platform using Jetson Orin AGX / NX.
Integrate Ouster OS1-32 LiDAR and VN100 IMU sensors for precise localization and motion control.
Implement real-time sensor data acquisition and synchronization for SLAM processing.

Voxel Coordinate Matching and Loop Closure Detection
Align LiDAR point cloud data with voxel-based global maps for real-time mapping and localization.
Develop loop closure detection algorithms that compare real-time LiDAR scans with global 3D voxel maps to correct accumulated drift.

Collision Avoidance and Autonomous Navigation Algorithm
Detect and track dynamic obstacles using voxel-based spatial computation.
Estimate size, velocity, and trajectory of obstacles for collision-free path planning.
Apply the developed SLAM module for autonomous mission execution in GPS-denied environments.

Data Optimization and Sensor Fusion
Store and compress high-resolution spatial data within voxel bounding boxes.
Fuse multiple sensor inputs (e.g., camera, thermal sensors) to enhance environmental perception.

Acknowledgement

• 국문표기: 

• English: 

Images:
Currently there are no images yet.