QLIO

Quantized LiDAR-Inertial Odometry

This work introduces a bandwidth-efficient, quantization-based LIO framework.

---

🚀 Key Features

• Distributed Quantized Architecture

  • Dual-processor pipeline with host-co-processor collaboration
  • 14.1× residual data compression through quantization

• Adaptive Resampling

  • rQ-vector-based feature selection (85% redundancy reduction)
  • Bandwidth-aware point cloud downsampling

• QMAP

  • Quantized-based MAP state estimation

---

📦 Installation

Very thanks for Fastlio! please download the MCD-ntu dataset and then:

mkdir catkin_ws/src
cd catkin_ws/src
# Clone repository
git clone https://github.com/luobodan/QLIO.git
# Build with catkin
cd ../
catkin_make
source devel/setup.bash
roslaunch fastlio mapping_ouster128_MCD_ntu.launch
rosrun your_dataset_path/*.bag