akshattrivedi / automating-traffic-signals-based-on-traffic-density-estimation-using-yolo

Automating the traffic signal timings using images of vehicles near the crossroads using YOLO (You Only Look Once) which includes Convolutional and Fully Connected Neural Networks which is implemented on Python-Django Web Framework.

Automating-Traffic-Signals-Based-on-Traffic-Density-Estimation-using-YOLO

Automating the traffic signal timings using images of vehicles near the crossroads with traffic signals using YOLO (You Only Look Once) and implemented on Python-Django Web Framework.

Usage

• Download YOLOv2 Weights: http://bit.ly/YOLOv2Weights
• Copy 'yolov2.weights' to 'bin/' folder
• Setup YOLO

$ python3 setup.py build_ext --inplace

• Run YOLO for image directory

$ ./flow --model cfg/run/yolo.cfg --load bin/yolov2.weights --imgdir sample_img/

• Retrieve JSON Values of objects inside the image with labels alongwith dimensions of the bounding box and confidence values.

$ python3 JSONValues.py

• HELP with YOLO options

$ ./flow --help

• Run the YOLO model on the Python-DJANGO website

$ python3 manage.py runserver

• There are two methods to run YOLO. Open Browser Type: localhost:8000

• Upload any image on index page and find out number of vehicles belonging to different classes like Car, Motorbike, Truck, Bus, Bicycle

• Run Traffic Simulation for an interval of 10 seconds by using the button 'Traffic Simulation' button on the index page.