4_4

Basic point cloud processing

• Load a lidar point cloud datum from KITTI dataset
• Use PCL functions to perform 1. Point cloud downsampling, 2. KD-tree search, 3. Passthrough filter, 4. Statistical outlier removal, 5. Plane detection

---

How to build & run

Requirement: PCL

Local build

mkdir build && cd build
cmake ..
make -j
cd ..
./build/downsampling
./build/kdtree
./build/passthrough
./build/sor
./build/plane_det

Docker build

Requires base build

docker build . -t slam:4_4
docker run -it --env DISPLAY=$DISPLAY -v /tmp/.X11-unix/:/tmp/.X11-unix:ro slam:4_4

# Inside docker container
cd fastcampus_slam_codes/4_4
./build/downsampling
./build/kdtree
./build/passthrough
./build/sor
./build/plane_det

---

Output

Passthrough filter (Remove LiDAR points from vehicle)

KD-Tree (Find nearest neighbor LiDAR points)

Downsampling

Floor plane detection

Statistical Outlier Removal (SOR)