ilovax/distancing_detector

Social-distancing-detector

Two interfaces: - Web interface - Terminal interface (using raspberry pi)

1-INTRODUCTION:

In this project we impliment a social distancing detector based on YOLO model for object detaction and split it into two parts:

1.1- Using web interface to upload a video and detect social distancing on it:

Using flask framework and a smiple web page to upload a video and watch the results.

1.2- Using CLI :

We can provide both the name of the input video and the output and must provide the yolo folder path.

2-Requirements:

• Requirements.txt contains the needed python libraries.

$ pip install -r requirements.txt 

• Download the yolo3 file:

coco.names

yolov3.cfg

yolov3.weights

• Move them to a folder

$ mkdir yolov3/
$ mv coco.name yolov3.cfg yolov3.weights yolov3/

3-Running the project

3.1 web server

$ flask run

Then access localhost:5000/

3.2 Cli

$ python distancing_detector.py input_video output_video yolo_path

yolo_path is the path to the directory with the yolov3 files (yolov3/). you can enter 'cam' instead of the input_video for capturing from camera