Tutorial: Feature Matching Using OpenCV in Python

Feature matching is a key task in computer vision for finding correspondences between features of two images. OpenCV provides robust tools for feature detection, description, and matching.

This tutorial will cover feature matching with different algorithms and practical examples.

What You’ll Learn

1. Introduction to Feature Matching

Feature matching involves:

1. Feature Detection: Identifying keypoints in images.
2. Feature Description: Computing descriptors for these keypoints.
3. Matching: Finding correspondences between descriptors from two images.

OpenCV provides:

• Brute-Force Matcher (cv2.BFMatcher)
• FLANN Matcher (cv2.FlannBasedMatcher)

2. Brute-Force Matching with ORB Features

ORB (Oriented FAST and Rotated BRIEF) is a fast and efficient feature detector/descriptor suitable for real-time applications.

Example: Brute-Force Matching with ORB

import cv2

# Load two images
image1 = cv2.imread("image1.jpg", cv2.IMREAD_GRAYSCALE)
image2 = cv2.imread("image2.jpg", cv2.IMREAD_GRAYSCALE)

# Initialize ORB detector
orb = cv2.ORB_create()

# Detect and compute features
kp1, des1 = orb.detectAndCompute(image1, None)
kp2, des2 = orb.detectAndCompute(image2, None)

# Brute-Force Matcher with Hamming distance
bf = cv2.BFMatcher(cv2.NORM_HAMMING, crossCheck=True)

# Match descriptors
matches = bf.match(des1, des2)

# Sort matches by distance
matches = sorted(matches, key=lambda x: x.distance)

# Draw matches
result = cv2.drawMatches(image1, kp1, image2, kp2, matches[:10], None, flags=cv2.DrawMatchesFlags_NOT_DRAW_SINGLE_POINTS)

# Display result
cv2.imshow("ORB Feature Matching", result)
cv2.waitKey(0)
cv2.destroyAllWindows()

3. FLANN-Based Matching

FLANN (Fast Library for Approximate Nearest Neighbors) is a fast matching algorithm suitable for large datasets.

Example: FLANN Matching with SIFT

import cv2

# Load two images
image1 = cv2.imread("image1.jpg", cv2.IMREAD_GRAYSCALE)
image2 = cv2.imread("image2.jpg", cv2.IMREAD_GRAYSCALE)

# Initialize SIFT detector
sift = cv2.SIFT_create()

# Detect and compute features
kp1, des1 = sift.detectAndCompute(image1, None)
kp2, des2 = sift.detectAndCompute(image2, None)

# Define FLANN matcher parameters
FLANN_INDEX_KDTREE = 1
index_params = dict(algorithm=FLANN_INDEX_KDTREE, trees=5)
search_params = dict(checks=50)

flann = cv2.FlannBasedMatcher(index_params, search_params)

# Match descriptors
matches = flann.knnMatch(des1, des2, k=2)

# Apply Lowe's ratio test
good_matches = []
for m, n in matches:
    if m.distance < 0.7 * n.distance:
        good_matches.append(m)

# Draw matches
result = cv2.drawMatches(image1, kp1, image2, kp2, good_matches, None, flags=cv2.DrawMatchesFlags_NOT_DRAW_SINGLE_POINTS)

# Display result
cv2.imshow("FLANN Feature Matching", result)
cv2.waitKey(0)
cv2.destroyAllWindows()

4. Using SIFT Features

SIFT (Scale-Invariant Feature Transform) is a robust algorithm for detecting and describing features invariant to scale and rotation.

Example: Brute-Force Matching with SIFT

import cv2

# Load two images
image1 = cv2.imread("image1.jpg", cv2.IMREAD_GRAYSCALE)
image2 = cv2.imread("image2.jpg", cv2.IMREAD_GRAYSCALE)

# Initialize SIFT detector
sift = cv2.SIFT_create()

# Detect and compute features
kp1, des1 = sift.detectAndCompute(image1, None)
kp2, des2 = sift.detectAndCompute(image2, None)

# Brute-Force Matcher
bf = cv2.BFMatcher(cv2.NORM_L2, crossCheck=True)

# Match descriptors
matches = bf.match(des1, des2)

# Sort matches by distance
matches = sorted(matches, key=lambda x: x.distance)

# Draw matches
result = cv2.drawMatches(image1, kp1, image2, kp2, matches[:10], None, flags=cv2.DrawMatchesFlags_NOT_DRAW_SINGLE_POINTS)

# Display result
cv2.imshow("SIFT Feature Matching", result)
cv2.waitKey(0)
cv2.destroyAllWindows()

5. Homography Estimation

After feature matching, you can compute the homography matrix to align two images or perform perspective transformation.

Example: Homography with SIFT Features

import cv2
import numpy as np

# Load two images
image1 = cv2.imread("image1.jpg", cv2.IMREAD_GRAYSCALE)
image2 = cv2.imread("image2.jpg", cv2.IMREAD_GRAYSCALE)

# Initialize SIFT detector
sift = cv2.SIFT_create()

# Detect and compute features
kp1, des1 = sift.detectAndCompute(image1, None)
kp2, des2 = sift.detectAndCompute(image2, None)

# Brute-Force Matcher
bf = cv2.BFMatcher(cv2.NORM_L2)
matches = bf.knnMatch(des1, des2, k=2)

# Apply Lowe's ratio test
good_matches = []
for m, n in matches:
    if m.distance < 0.7 * n.distance:
        good_matches.append(m)

# Extract points
src_pts = np.float32([kp1[m.queryIdx].pt for m in good_matches]).reshape(-1, 1, 2)
dst_pts = np.float32([kp2[m.trainIdx].pt for m in good_matches]).reshape(-1, 1, 2)

# Find homography
H, mask = cv2.findHomography(src_pts, dst_pts, cv2.RANSAC, 5.0)

# Warp perspective
height, width = image2.shape
aligned_image = cv2.warpPerspective(image1, H, (width, height))

# Display result
cv2.imshow("Aligned Image", aligned_image)
cv2.waitKey(0)
cv2.destroyAllWindows()

6. Practical Examples

6.1 Real-Time Feature Matching with ORB

import cv2

# Initialize ORB detector
orb = cv2.ORB_create()

# Load the template image
template = cv2.imread("template.jpg", cv2.IMREAD_GRAYSCALE)
kp_template, des_template = orb.detectAndCompute(template, None)

# Open webcam
cap = cv2.VideoCapture(0)

while True:
    ret, frame = cap.read()
    if not ret:
        break

    # Convert to grayscale
    gray_frame = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)

    # Detect and compute features
    kp_frame, des_frame = orb.detectAndCompute(gray_frame, None)

    # Match features
    bf = cv2.BFMatcher(cv2.NORM_HAMMING, crossCheck=True)
    matches = bf.match(des_template, des_frame)
    matches = sorted(matches, key=lambda x: x.distance)

    # Draw matches
    result = cv2.drawMatches(template, kp_template, frame, kp_frame, matches[:10], None, flags=2)

    # Display result
    cv2.imshow("Real-Time Feature Matching", result)

    if cv2.waitKey(1) & 0xFF == ord('q'):
        break

cap.release()
cv2.destroyAllWindows()

6.2 Draw Keypoints on Images

import cv2

# Load an image
image = cv2.imread("image.jpg", cv2.IMREAD_GRAYSCALE)

# Initialize ORB detector
orb = cv2.ORB_create()

# Detect keypoints
keypoints = orb.detect(image, None)

# Draw keypoints
result = cv2.drawKeypoints(image, keypoints, None, color=(0, 255, 0), flags=0)

# Display result
cv2.imshow("Keypoints", result)
cv2.waitKey(0)
cv2.destroyAllWindows()

7. Summary

Key Functions

• cv2.BFMatcher: Brute-force matcher for descriptor matching.
• cv2.FlannBasedMatcher: Fast matcher for large datasets.
• cv2.findHomography: Compute homography matrix for image alignment.
• cv2.drawMatches: Visualize matches between two images.

Best Practices

1. Use Lowe’s ratio test to filter good matches.
2. Select feature detectors (e.g., ORB, SIFT) based on task requirements.
3. Use homography for perspective transformations after matching.