- Image Recognition using SIFT and BRISK feature detectors
- We are using the latest
opencvlibrary version3.0.0rc1 - Run
pip install -r requirements.txt - Test running it by
python beyondbagsoffeatures.py
- Scipy
- Numpy
- scikit-learn
- matplotlib
- The image data set need to run the code is provided in the
imagesfolder. (User may change it as per requirement.) - Currently there are 3 datasets namely
trainingandtestingform the 1st paircaltech_trainandcaltect_testform the 2nd pairc1_trainandc1_testform the last pair
- User can change the dataset one wants to run by commenting the
line#30-37ofbeyondbagsoffeatures.py. - User can switch between SIFT and BRISK by setting
sift=Falseonline#28
- S. Lazebnik, C. Schmid, and J. Ponce, “Beyond Bags of Features: Spatial Pyramid Matching for Recognizing Natural Scene Categories”, IEEE CVPR, 2006
- https://github.com/wihoho/Image-Recognition