킥라니 멈춰! BE

FE : https://github.com/SangRakee/stop_kickrani_FE

Jetson-nano : https://github.com/bururaru/stop_kickrani_jetsonnano

Table of Contents

1. About The Project
   • Used Deivce
   • Built With
   • Architecture
2. Trained Indicators
3. Getting Started
   • Installation
4. Usage example
5. Roadmap
6. Contact

About The Project

This is the project detecting illegal electric scooter riders on the public street with unmanned device. And this repository contains the 2nd part of the detection (number of person/helmet, kickboard brand). It receives the rider data from jetson nano and detects the object with YOLOv5 model and sends the data to Database.

Used Device

• AWS EC2 with GPU
• GPU : NVIDIA M60 8GB

Built With

• Python 3.8.10

• Django 3.2.3

• pytorch 1.8.1

• YOLOv5

• MQTT

Architecture

Trained Indicators

Model	number of dataset	mAPval 0.5	mAPtest 0.5
Rider (x)	983	0.981	0.955
Kickboard (m)	1500	0.94	0.843
Person (s)	712	0.992	0.978
helmet (m)	629	0.984	-

Getting Started

To get a local copy up and running follow these simple steps.

Installation

1. Clone the repo

   git clone https://github.com/bururaru/stop_kickrani_BE.git

2. Install libraries with requirements.txt

   pip install -r requirements.txt

3. Run the Django server

Usage example

Dashboard

Roadmap

See the open issues for a list of proposed features (and known issues).

Contact

김기범 - kimgibum90@gmail.com

이상락 - leesangrak0307@gmail.com

이창민 - bururaru@gmail.com

장소영 - so970404@gmail.com

정수경 - skjung247@gmail.com

Project Link: https://github.com/bururaru/stop_kickrani_BE