Skip to content

Akhtar303/Weapon-Detection-in-CCTV-Footage

 
 

Repository files navigation

Weapon-Detection-in-CCTV-Footage

Repository contains implementation of weapon detection pipeline which consists of 2 independently trained object detection and one classification algorithm.

VGG_16 was trained using PyimageSearch tutorial on gun(~11000) images.

Requirements

Python 3.6 or later

  • numpy
  • opencv-python
  • torch >= 1.1.0

Inference

VGG

images : python3 predict_VGG.py --image test_imgs/gun.jpg

webcame : python3 webcam_VGG.py

Pre-Trained Model Weights

YOLO Backend weights.

About

Computer vision pipeline to detect weapon in surveillance videos

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 78.9%
  • Python 19.6%
  • Shell 1.2%
  • Dockerfile 0.3%