AI-Powered Surveillance Vehicle

Overview

This project is an AI-powered surveillance vehicle equipped with a 360° movable camera, enabling real-time control and object detection. The system leverages ESP32-CAM for movement and streaming, YOLO for object detection, and a PyQt-based desktop application for control and monitoring. The vehicle is controlled through custom API routes to manage motors, servos, lights, and stream the camera feed.

Features

• 360° mobility for the surveillance vehicle
• Real-time video streaming at 720p
• YOLO-based object detection for identifying objects in the stream
• Control via web API (POSTMAN tested)
• PyQt desktop application with multi-threading for smooth control and monitoring

API Endpoints

Motor Control:

• http://192.168.137.234/right_motors?control=0&value=80
  Controls the right motor.

  • control=0: Stop/reset
  • control=1: Move forward
  • control=-1: Move backward
  • value: Speed level (0-100)

• http://192.168.137.234/left_motors?control=1&value=100
  Controls the left motor.

  • control=0: Stop/reset
  • control=1: Move forward
  • control=-1: Move backward
  • value: Speed level (0-100)

Servo Motors:

• http://192.168.137.234/servo_motors?control=1&value=90
  Controls the servo motors.
  • control=1: Move servo
  • value: Degree adjustment (±90 degrees from the last position)

LED Control:

• http://192.168.137.234/led_blink?control=2&value=800
  Controls the LED blinking.
  • control=2: Blink
  • value: Frequency of blinking (in ms)

Flash Light:

• http://192.168.137.234/flash_light?control=1&value=85
  Controls the flash light.
  • control=1: Turn on
  • value: Brightness level (0-100)

Video Stream:

• http://192.168.137.234:81/stream
  Streams the live video feed from the camera.

Hardware

• ESP32-CAM: Microcontroller with a built-in camera used for live video streaming.
• Motors and Servos: For vehicle movement and camera angle adjustments.
• LEDs: Used for illuminating the surroundings.

Software

• Firmware: Developed using C/C++ on the ESP32 using the Arduino IDE.
• Backend: Custom web server created with API endpoints for vehicle control and interaction.
• Desktop Application: Built with PyQt, handling real-time control, video stream, and YOLO object detection.
• Object Detection: Implemented using YOLO for detecting objects in real-time within the video feed.

Tech Stack

• Firmware: C/C++
• Software: Web API, API Development, Postman, Multi-threading
• AI: YOLO, Object Detection
• UI: PyQt, UI/UX

Installation

Firmware Installation

1. Download and install the Arduino IDE.
2. Install the ESP32 board manager in Arduino IDE.
3. Upload the ESP32-CAM code to the device via USB.
4. Connect the vehicle’s hardware components (motors, LEDs, servos) to the appropriate GPIO pins.

Desktop Application Installation

1. Clone this repository:

      git clone https://github.com/your-username/AI-Powered-Surveillance-Vehicle.git
      cd AI-Powered-Surveillance-Vehicle

2. Install dependencies:

      For Pyhon Environment : Numpy, OpenCV, PyQt5, Ultralytics, YOLOv10 model etc.
      For Arduino IDE:        AsyncTCP, ESP Async WebServer, ESP32Servo etc.

3. Start the ESP32 vehicle and the pyqt desktop application.
4. Get the IP address of ESP32 from your router/ phone/ Laptop.
5. Enter the IP address in the PyQt Desktop Application.

Usage

• Control the vehicle using the desktop application.
• Use the provided API routes to interact with the vehicle remotely, such as controlling motors, LEDs, or servo positions.

Future Enhancements

• Improved object detection models for better accuracy and speed.
• Integration with cloud services for remote control.
• Voice-based control interface.

Contributors

• Tamal Mallick : Firmware coding, Software coding, Machine Learning Integration, Wiring, Project Planning, Report & PPT making, Background Research, Others
• Avishek Mondal : 3D Printing, Circuit Design, Hardware Assembly, Wiring, Project Planning, PPT making, Background Research, Others
• Souvik Baidya : Project Planning, Co-ordination, Report & PPT making, Background Research, Others