Two interfaces: - Web interface - Terminal interface (using raspberry pi)
In this project we impliment a social distancing detector based on YOLO model for object detaction and split it into two parts:
Using flask framework and a smiple web page to upload a video and watch the results.
We can provide both the name of the input video and the output and must provide the yolo folder path.
- Requirements.txt contains the needed python libraries.
$ pip install -r requirements.txt
- Download the yolo3 file:
- Move them to a folder
$ mkdir yolov3/
$ mv coco.name yolov3.cfg yolov3.weights yolov3/
$ flask run
Then access localhost:5000/
$ 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