The main goal of this project is to develop an automated visual surveillance system (AVSS) consisting of a motion detection and tracking system among the various components.
The AVSS must eventually satisfy the following characteristics:
- work in real-time
- work with moving cameras: it should be able to work e.g. with PTZ cameras
- work indoor and outdoor: system must be robust to changing lighting conditions
Right now, only a simple Python-based motion detection and tracking system has been implemented using two basic background substitution models:
- First background model: first frame taken to model the background
- Second background model: weighted average of frames
The Python code implementation for these two background models are based from Adrian Rosebrock's tutorials @ www.pyimagesearch.com
You can find more information about this simple motion detection system @ README.md There you will find instructions on how to install and use the system.
First frame background model: only the first 300 first frames are shown, @ 4x normal speed
Weighted average background model: only the first 300 first frames are shown, @ 4x normal speed