Crowd Monitoring: Abnormal Behaviour Detection (FINISHED! Hooray!)
This project applied computer vision and pattern recognition methods aimed to detect abnormal behaved object in crowd, such as biker (fast motion) in a crowd of walking (slow motion) people.
- Language: Python3.6.1
- Dependencies: OpenCV, Scikit-Learn, Scikit-Image, and Numpy (Anaconda recommanded)
- Feature extraction: Morphological filter, Normalisation, Optical Flow
- Classification Model: SVM, LogisticRegression, KNN
Please run ./code/Core.py for demo
or check the ./results folder for videos and images