videoanalytics

This project is about videosources emergency detection made with ZeroMQ infrastructure

Used technologies

• ZeroMQ
• YOLOv8
• PSNR
• OpenCV
• Multiprocessing & Threading

Prerequisites

• Torch with CUDA
• Python3.12

Preparations

• pip install -r requirements.txt
• Download all necessary files from Google Drive :
• (optional) place test videos & meanframes to ./videos folder
• place YOLO models to ./models folder
• set up config.json with actual files locations, ip addresses & ports, psnr threshold

Notes:

meanframes must have same resolution as original videos, '.jpg' hardcoded
meanframes were created on "normal" parts of videos with psnr.py

Meanframes for PSNR on custom videos

Create meanframes for PSNR function:

• Prepare "normal" videofragment (not containing emergency situations) in video editor
• Use psnr.py:
  • python psnr.py <videopath>
• Meanframe will be created in same directory with same filename in '.jpg' format

Usage

After you've set up config.json, just:

python run.py

Ctrl+C to stop

TODOs

• create meanframes automatically
• train models on another datasets
• add resolution to config
• Ctrl+C handler
• Auto scale processes
• Test on realtime videosources

Example