Number Plate Detection Webpage

! Number Plate recognition web page

Tech Stack

Computer Vision

• OpenCV: Used for opening the camera, capturing frames, converting images to grayscale, drawing rectangles, and displaying text on images.
• Haar Cascade Classifier: Specifically, haarcascade_russian_plate_number.xml for detecting number plates.

OCR (Optical Character Recognition)

• Pytesseract: Used for extracting text from the detected number plates.
  • Parameter Tuning:

    custom_config = r'--oem 3 --psm 1'

  • OCR Engine Mode (--oem):
    • 0: Legacy engine only.
    • 1: Neural nets LSTM engine only.
    • 2: Legacy + LSTM engines.
    • 3: Default, based on what is available.
  • Page Segmentation Mode (--psm):
    • 0: Orientation and script detection (OSD) only.
    • 1: Automatic page segmentation with OSD.
    • 2: Automatic page segmentation, but no OSD, or OCR.
    • 3: Fully automatic page segmentation, but no OSD.
    • 4: Assume a single column of text of variable sizes.
    • 5: Assume a single uniform block of vertically aligned text.
    • 6: Assume a single uniform block of text.
    • 7: Treat the image as a single text line.
    • 8: Treat the image as a single word.
    • 9: Treat the image as a single word in a circle.
    • 10: Treat the image as a single character.
    • 11: Sparse text. Find as much text as possible in no particular order.
    • 12: Sparse text with OSD.
    • 13: Raw line. Treat the image as a single text line, bypassing hacks that are Tesseract-specific.

Text Filtering

• Regular Expressions (re): Used to filter out non-alphanumeric characters from the recognized text.

Web Application

• Flask: Used for web application framework to render templates, jsonify responses, and handle requests.
  • Key functions: render_template, jsonify, Response, request.

Data Storage

• CSV Library: Used to append detected number plates to a CSV file for record-keeping.

Data Display

• Pandas: Used for loading and displaying data on the webpage after being converted to JSON format.

Interrupt Handling

• Signal: Used to send a SIGINT signal to interrupt the current Python process.

Timing Control

• Time: Used for controlling the frame capture timing until a number plate is recognized.

Implementation Logic

1. Camera Initialization and Frame Capture:

   • The camera opens and starts capturing frames.
   • Each captured frame is converted to grayscale for processing.

   import cv2
   
   cap = cv2.VideoCapture(0)
   while True:
       ret, frame = cap.read()
       gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
       # Further processing

2. Number Plate Detection:

   • The grayscale frame is sent to the recognition engine using the Haar Cascade Classifier to detect number plates.
   • If a number plate is detected, a rectangle is drawn around it, and a label is put with the recognized text.
   • If no number plate is detected, no label is put.

   plate_cascade = cv2.CascadeClassifier('haarcascade_russian_plate_number.xml')
   plates = plate_cascade.detectMultiScale(gray, scaleFactor=1.1, minNeighbors=4)
   
   for (x, y, w, h) in plates:
       cv2.rectangle(frame, (x, y), (x + w, y + h), (0, 220, 0), 1)
       # Further processing in source file

3. Displaying on screen:

   Response(generate_frames(), mimetype='multipart/x-mixed-replace; boundary=frame')

4. Saving to CSV

5. Interrupt Handling

6. Timing Control:

   • The time module is used to pause the frame capture until a number plate is detected. =======

Car-Plate-detection

Computer Vision Implementation with PyTesseract and Haar Cascade This repository features a computer vision project using PyTesseract and Haar Cascade to detect number regions and recognize characters. Built with Flask, it provides a web interface for user interaction. Detected numbers are stored in a CSV file with timestamps.