Trinetra

- Let us be protected by the third eye!

Automatic Number Plate Detection Recognition and Traffic Light Violation Detection

Here, I have not provided the model that I have trained so that nobody can misuse it. If you want the model, can contact me at LinkedIn or Email

Video Demonstration(Youtube)

Automatic Number Plate Recognition

alpr.mp4

Traffic Light Violation

trafficlight.mp4

Overview

This is the repository for the Trinetra project. Trinetra is a project that aims to provide a solution for automatic number plate recognition and the traffic light violation for the nepali number plates. The project is divided into three parts: the backend, the frontend, and the services. The backend is responsible for the database and the API. The frontend is responsible for the user interface. The services are responsible for the machine learning models that are used to detect vehicles, number plates, speed, and traffic lights.

Objectives

• To Automate the process of number plate detection and recognition.
• To Automate the process of traffic light violation detection.
• To Monitor the recognition of the number plate and traffic light violation detection in real time and store the data. in the database for future reference and further analysis.
• To provide a user-friendly interface for the user to interact with the system.
• To monitor the images of number plate recognitions and traffic light violations in real time.

Features

• Number Plate Detection and Recognition
• Traffic Light Violation Detection
• Real-time monitoring of number plate recognition and traffic light violation detection
• user-friendly interface for the user to interact with the system.
• Search and filter the number plate recognition and traffic light violation detection data.
• Monitor the images of number plate recognitions and traffic light violations in real time.

Technologies Used

For the backend, the following technologies are used:

• Python
• Django
• Django Rest Framework
• PostgreSQL
• JWT Authentication

For the frontend, the following technologies are used:

• React
• Material UI
• Axios
• Redux
• Different Chart Libraries

For the services, the following technologies are used:

• Python
• OpenCV
• YOLO
• OCR
• OOPs Concept
• Different Machine Learning Libraries
• CNN

Services for Trinetra

Click to expand

1. Number Plate Detection and Recognition

• Number plate detection and recognition is the process of detecting the number plate of a vehicle and recognizing the characters on the number plate. The process involves the following steps:

  • Number Plate Detection

  • Number Plate Recognition

  • ALPR

2. Traffic Light Violation Detection

• Traffic light violation detection is the process of detecting the violation of traffic lights by vehicles. The process involves the following steps:

  • Traffic Light Detection

  • Vehicle Detection

  • Traffic Light Violation Detection

  • Traffic light violation

How to Install

Click to expand

1. Clone the repository

git clone https://github.com/ishworrsubedii/automatic_number_plate_detection_recognition-traffic_light_violation.git

2. Create Environment

conda create -n trinetra python=3.10

3. Install the requirements

pip install -r requirement

Backend

For the backend, open the suitable IDE, i.e., Pycharm, and set up the Conda environment.

Frontend

For frontend, open the suitable idea, i.e., vs. code, and open the frontend-trinetra directory.The requirements for the frontend are:

• Node package manager needs to be installed in the system

npm install

To start the frontend, run the following command:

npm start 

To build the optimized version of the frontend, run the following command:

npm run build

Services

1. Setup

cd services_trinetra

setup.py install

Copy the package of thelib and add it into the site-packages of the python environment

For alpr

Change the IP of ipcam inside backend_trinetra/alpr/alprservices.py to your IP.

For traffic lights

Change the IP of the ipcam inside backend_trinetra/trafficlight/trafficlightservices.py.

Contributors

Contributors to the project are always welcome. You can contribute to the project by forking the repository and creating a pull request. You can also create an issue if you find any bugs or have any suggestions for the project.