Traffic Violation Detection

Real-time Speed violation detection and Lane-Line violation detection using Yolov4 and Deep-Sort

Instructions

First clone the repository
Run the conda-env.yml to install a dedicated conda environment and install all the necessary repositories.

conda env create -f conda-env.yml
conda activate yolov4-gpu

If you don't have a gpu, comment out 8 and 9 lines and remove '-gpu' segment in 11 th line. (tensorflow==2.3.0 instead of tensorflow-gpu==2.3.0)
Please note that you need a comparatively powerful gpu in order to get a descent result.
Download yolov4 pre-trained weights --> https://github.com/AlexeyAB/darknet/releases/download/darknet_yolo_v3_optimal/yolov4.weights
Move the yolov4.weights file to /data folder
Excute following command to convert darknet weights to tensorflow model.

python save_model.py --model yolov4

python object_tracker_speed_violation.py --output ./outputs/processed_vids/speed.avi --model yolov4

Run this commad to test the Lane-Line violation detection program after changing the paths in lane_line_extract3.py (5th line, change according to your PC path)

python lane_line_viol_detection.py --output ./outputs/processed_vids/lane.avi --model yolov4

python parking_violation_detection.py  --output ./outputs/processed_vids/parking.avi --model yolov4