Multi Object Multi Camera Tracking for Traffic Surveillance
The main MultiVideoProcessor only requires the following dependencies:
- Python >= 3.6
- Pytorch >= 1.0
For all detector/tracker/ReID algorithm, please refer to their respective Github page for their requirements.
- Open config.ini and update entries accordingly, key entries to take note of are as follows:
- query_video
- gallery_video_folder
- det_algorithm
- reid_algorithm
- Ensure all weights and model path are correct for their respective algorithm
- Activate conda environment
- Run the MultiVideoProcessor with the follow command
python MultiVideoProcessor.py
Please refer to add_algorithm.md for detailed step by step guide.