This repo functions mostly the same with this, but is less heavy. It does not rely on Boost, I have re-written most of the code to utilise the power of pybind11.
import pydbow3 as bow
voc = bow.Vocabulary()
voc.load("ORBvoc.txt")
db = bow.Database()
db.setVocabulary(voc)
# extract features using OpenCV
...
# add features to database
for features in features_list:
db.add(features)
# query features
feature_to_query = 1
results = db.query(features_list[feature_to_query])- OpenCV
- CMake
- DBoW3
- clone the repo with
--recursive - build and install
modules/DBoW3 - install with
pip install . - currently only
*nixis supported
This work is based on https://github.com/foxis/pyDBoW3 and https://github.com/edmBernard/ybind11_opencv_numpy.git