Document Scanner

Scan documents using edge detection and perspective transformation.

View Source Code

Overview

Transform photos of documents into clean, flat scans - like a mobile scanning app.

Key Techniques:

  • Canny edge detection
  • Contour detection and approximation
  • Perspective transformation
  • Adaptive thresholding

How It Works 

Input Image → Edge Detection → Find Document → Perspective Warp → Output
     ↓              ↓               ↓                ↓
 [Photo of    [Canny edges]  [4-corner        [Flattened
  document]                   contour]         document]

Pipeline Steps

  1. Preprocessing: Convert to grayscale, apply Gaussian blur
  2. Edge Detection: Canny edge detector finds edges
  3. Contour Finding: Find largest 4-sided contour (the document)
  4. Corner Ordering: Sort corners to top-left, top-right, bottom-right, bottom-left
  5. Perspective Transform: Warp to bird’s-eye view
  6. Enhancement: Adaptive threshold for clean text

Key OpenCV Functions 

# Edge detection
edges = cv2.Canny(blur, 50, 200)

# Find contours
contours, _ = cv2.findContours(edges, cv2.RETR_LIST, cv2.CHAIN_APPROX_SIMPLE)

# Approximate to polygon
approx = cv2.approxPolyDP(contour, epsilon, True)

# Perspective transform
M = cv2.getPerspectiveTransform(src_pts, dst_pts)
warped = cv2.warpPerspective(image, M, (width, height))

# Enhance text
result = cv2.adaptiveThreshold(warped, 255,
    cv2.ADAPTIVE_THRESH_GAUSSIAN_C, cv2.THRESH_BINARY, 11, 2)

Controls

Key Action
c Capture and process document
t Toggle threshold enhancement
s Save scanned document
r Reset
q Quit

Running the Application 

python curriculum/applications/01_document_scanner.py

Official Documentation