Automotive

Automotive is an innovative project that combines machine learning, image processing, and MERN stack technologies to provide automated solutions for vehicle-related tasks such as number plate recognition, parking space detection, and vehicle counting.

Features

1. Number Plate Recognition

• Objective: Recognize vehicle number plates from video.
• Technology: Implemented using YOLOv8 for object detection and EasyOCR for text extraction.
• Key Points:
  • Uses two YOLO models: YOLOv8 for vehicle detection and a custom-trained YOLO model for detecting number plates from vehicles.
  • SORT (Simple Online Realtime Tracking) for object tracking.

2. Parking Space Detection

• Objective: Detect whether parking spaces are available or not.
• Technology: Image processing using OpenCV and cvzone.
• Key Points:
  • Detects empty parking spaces by counting white dots on binary-processed images.
  • Edge filtering and thresholding techniques are applied to determine availability.

3. Vehicle Counting

• Objective: Count vehicles in a video stream.
• Technology: YOLOv8 for vehicle detection and SORT for tracking.
• Key Points:
  • Detects vehicles in real-time and counts them in the video.
  • Generates output video with tracked and counted vehicles.

Tech Stack

Machine Learning & Image Processing

• YOLOv8: Used for vehicle and number plate detection.
• SORT: For object tracking.
• EasyOCR: For extracting text from detected number plates.
• OpenCV: For video capture and image manipulation.
• Numpy, Pandas: For data manipulation.

MERN Stack

• Frontend: Implemented using React.js.
• Backend: Node.js with Express.js.
• Database: MongoDB for storing user history.
• Authentication: Clerk authentication.
• Cloud Storage: Cloudinary for storing video history.