VISION Gate – End-to-End IoT & AI Smart Parking Solution

---

⭐️ If you like this project, don’t forget to give it a **star** on GitHub! ⭐️

---

Overview

VISION Gate is a smart parking system where users register via the mobile app.
When approaching the garage, the camera scans the license plate and checks it against the system for secure, automated entry.

---

System Architecture

Upcoming System Update (Expected)

---

Backend & Hardware

• NestJS Backend with clean software architecture.
• AI Model Integration for license plate detection.
• Flutter Mobile Application for user interaction.
• ESP32 Firmware to manage sensors, servos, and LEDs.
• MQTT-based real-time communication between backend, ESP32, and mobile app.
• Custom camera detection app integrated with backend APIs.
• Dockerized backend, deployed on Azure Container Service with remote operation.

Backend on Azure:

---

Wiring & Electronics

Designed and connected circuits with sensors and actuators.

Components Used:

• LDR (Light Dependent Resistor): light sensing
• IR Sensor: object detection
• Buzzer: alerts/notifications
• Servo Motor: mechanical movement (garage gate control)
• LCD: status messages
• LEDs (x3): system feedback indicators

Circuit Prototype:

---

IoT & Communication

• Integrated MQTT for real-time communication.
• Implemented API Client for RESTful interaction with backend.
• Achieved seamless communication between ESP32 → Backend → Mobile App.

---

Software & Architecture

• Developed with Platform IO, structured in modules (peripherals as classes).
• Dependency Injection (DI): for flexibility and testability.
• Single Responsibility Principle (SRP): each module has one job.
• Produced clean, maintainable, reusable code.

API & Communication Demo:

---

Database Design

📌 Organized into Schemas for clarity and separation of concerns.

• Users Table → first_name, last_name, address, phone, DOB, user_id, email, image_link, password
• Cars Table → car_id, company, car_model, user_id, plan_id, subscription_start
• Plans Table → plan_id, value, unit, price

Database Diagrams:

---

AI Model

Objectives:

• Detect license plates from images.
• Extract plate numbers as text.
• Provide API for integration.

Tools & Tech:

• YOLO → plate detection
• OpenCV → image preprocessing
• EasyOCR → text recognition
• Flask API → serve predictions

Pipeline:

1. API receives image
2. YOLO detects bounding box
3. OpenCV crops plate
4. EasyOCR extracts text
5. Flask returns license number

Detection Examples:

---

Flutter Mobile Application

• Designed using FlutterFlow.
• Dark/Light Mode support starting from login.
• Built 11 pages with smooth navigation.
• Added animations to enhance UX.
• Integrated with backend APIs and tested thoroughly.

  screens

  
  
  
  
  
  
  
  
  
  
  
  

---

Testing & Extras

• Test scripts → test.sql
• Database dump and migrations included.

---

Summary

VISION Gate is a fully integrated IoT + AI + Cloud system, combining:

• IoT firmware (ESP32 + sensors + actuators)
• Cloud backend (NestJS, Docker, Azure)
• AI-driven detection (YOLO + OpenCV + EasyOCR)
• Mobile App (Flutter)

A real-world, production-ready smart parking solution.

---