Template matching using OpenCV in Python

Template matching is a technique in computer vision used to find a sub-picture (template) in an image. It's a way to locate specific objects within the larger image. OpenCV provides a convenient function, cv2.matchTemplate(), for this purpose.

Here's a tutorial on template matching using OpenCV in Python:

1. Setup:

First, make sure you've installed OpenCV:

pip install opencv-python

Then, import the required libraries:

import cv2
import numpy as np

2. Read the Image and the Template:

main_image = cv2.imread('main_image.jpg', cv2.IMREAD_GRAYSCALE)
template = cv2.imread('template.jpg', cv2.IMREAD_GRAYSCALE)

In this example, both the main image and the template are loaded in grayscale, which simplifies the process. Adjust as needed.

3. Template Matching:

Use the cv2.matchTemplate() function:

result = cv2.matchTemplate(main_image, template, cv2.TM_CCOEFF_NORMED)

There are various methods available such as cv2.TM_CCOEFF, cv2.TM_CCOEFF_NORMED, cv2.TM_CCORR, etc. The method cv2.TM_CCOEFF_NORMED is used here as it usually provides good results, but you can experiment with others to see which works best for your specific use case.

4. Find the Best Match Location:

min_val, max_val, min_loc, max_loc = cv2.minMaxLoc(result)

The cv2.minMaxLoc() function returns the minimum and maximum pixel values and their locations. For cv2.TM_CCOEFF_NORMED, the maximum location (max_loc) corresponds to the top-left corner of where the template best matches the main image.

5. Draw a Rectangle around the Matched Area:

top_left = max_loc
bottom_right = (top_left[0] + template.shape[1], top_left[1] + template.shape[0])
cv2.rectangle(main_image, top_left, bottom_right, 255, 2)

Here, 255 is the color of the rectangle (white in this grayscale image) and 2 is the thickness of the rectangle's lines.

6. Display the Result:

cv2.imshow('Detected', main_image)
cv2.waitKey(0)
cv2.destroyAllWindows()

Complete Code:

Putting it all together:

import cv2
import numpy as np

# Read the images
main_image = cv2.imread('main_image.jpg', cv2.IMREAD_GRAYSCALE)
template = cv2.imread('template.jpg', cv2.IMREAD_GRAYSCALE)

# Template matching
result = cv2.matchTemplate(main_image, template, cv2.TM_CCOEFF_NORMED)
min_val, max_val, min_loc, max_loc = cv2.minMaxLoc(result)

# Draw a rectangle around the matched area
top_left = max_loc
bottom_right = (top_left[0] + template.shape[1], top_left[1] + template.shape[0])
cv2.rectangle(main_image, top_left, bottom_right, 255, 2)

# Display the result
cv2.imshow('Detected', main_image)
cv2.waitKey(0)
cv2.destroyAllWindows()

Run the script, and you should see the area where the template matches the main image highlighted with a rectangle. Remember that template matching is sensitive to size, rotation, and sometimes lighting changes, so it's most effective when the template is a direct subset of the main image.

Template matching with OpenCV in Python: Template matching is a technique in computer vision to find a template (small image) within a larger image. OpenCV provides a function called cv2.matchTemplate for template matching.

import cv2
import numpy as np

# Read the main image and the template
main_image = cv2.imread('main_image.jpg')
template = cv2.imread('template_image.jpg')

# Convert both images to grayscale
gray_main = cv2.cvtColor(main_image, cv2.COLOR_BGR2GRAY)
gray_template = cv2.cvtColor(template, cv2.COLOR_BGR2GRAY)

# Perform template matching
result = cv2.matchTemplate(gray_main, gray_template, cv2.TM_CCOEFF_NORMED)

# Get the coordinates of the best match
min_val, max_val, min_loc, max_loc = cv2.minMaxLoc(result)

# Draw a rectangle around the matched area
top_left = max_loc
bottom_right = (top_left[0] + template.shape[1], top_left[1] + template.shape[0])
cv2.rectangle(main_image, top_left, bottom_right, (0, 255, 0), 2)

# Display the result
cv2.imshow('Matched Result', main_image)
cv2.waitKey(0)
cv2.destroyAllWindows()

Real-time template matching in OpenCV: Template matching can be applied in real-time by continuously capturing frames from a camera.

import cv2
import numpy as np

# Open a connection to the webcam (0 represents the default camera)
cap = cv2.VideoCapture(0)

# Read the template image
template = cv2.imread('template_image.jpg')
gray_template = cv2.cvtColor(template, cv2.COLOR_BGR2GRAY)

while True:
    # Read a frame from the webcam
    ret, frame = cap.read()

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

    # Perform template matching
    result = cv2.matchTemplate(gray_frame, gray_template, cv2.TM_CCOEFF_NORMED)

    # Get the coordinates of the best match
    _, _, _, max_loc = cv2.minMaxLoc(result)

    # Draw a rectangle around the matched area
    top_left = max_loc
    bottom_right = (top_left[0] + template.shape[1], top_left[1] + template.shape[0])
    cv2.rectangle(frame, top_left, bottom_right, (0, 255, 0), 2)

    # Display the result
    cv2.imshow('Real-time Template Matching', frame)

    # Break the loop if 'q' is pressed
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break

# Release the VideoCapture object
cap.release()
cv2.destroyAllWindows()

Handling multiple matches in template matching with OpenCV: Use thresholding and non-maximum suppression to handle multiple matches.

import cv2
import numpy as np

main_image = cv2.imread('main_image.jpg')
template = cv2.imread('template_image.jpg')

gray_main = cv2.cvtColor(main_image, cv2.COLOR_BGR2GRAY)
gray_template = cv2.cvtColor(template, cv2.COLOR_BGR2GRAY)

# Perform template matching
result = cv2.matchTemplate(gray_main, gray_template, cv2.TM_CCOEFF_NORMED)

# Set a threshold for matches
threshold = 0.8
loc = np.where(result >= threshold)

# Draw rectangles around multiple matches
for pt in zip(*loc[::-1]):
    cv2.rectangle(main_image, pt, (pt[0] + template.shape[1], pt[1] + template.shape[0]), (0, 255, 0), 2)

# Display the result
cv2.imshow('Multiple Matches', main_image)
cv2.waitKey(0)
cv2.destroyAllWindows()