- uses KITTI dataset
- detects ORB features from 5 images
- Creates vocabulary tree using DBoW2
- Matches features from 5 images using DBoW2
Requirement: OpenCV4, DBoW2
(DBoW3, FBoW can be used as well - but API will be slightly different)
mkdir build && cd build
cmake ..
make -j
./bow_matching
Requires base build
docker build . -t slam:3_5
docker run -it --env DISPLAY=$DISPLAY -v /tmp/.X11-unix/:/tmp/.X11-unix:ro slam:3_5
# Inside docker container
cd fastcampus_slam_codes/3_5
./build/bow_matching