Home - News ( United States | United Kingdom | Italy | Germany ) - Football scores

Showing content from https://www.geeksforgeeks.org/python/python-opencv-canny-function/ below:

Python OpenCV - Canny() Function

Last Updated : 23 Jul, 2025

Canny edge detection algorithm is used in computer vision for identifying edges within an image. It helps in highlighting boundaries which are important for tasks like object detection and image segmentation. In this article, we will see how OpenCV's built-in Canny() function detects edges in an image.

Syntax: cv2.Canny(image, T_lower, T_upper, aperture_size, L2Gradient)

Where: 

• Image: Input image to which Canny filter will be applied.
• T_lower: Lower threshold value in Hysteresis Thresholding.
• T_upper: Upper threshold value in Hysteresis Thresholding.
• aperture_size: Aperture size of Sobel filter.
• L2Gradient: Boolean parameter used for more precision in calculating Edge Gradient.

Return Value: Returns an image with detected edges where edges are marked in white (255) and non-edges in black (0).

Input Image:

• cv2.Canny(img, t_lower, t_upper): Applies Canny edge detection algorithm on the image with the given lower and upper threshold values.
• t_lower = 50: Sets lower threshold for edge detection in the Canny algorithm.
• t_upper = 150: Sets upper threshold for edge detection in the Canny algorithm.
• cv2.imshow('original', img): Displays the original image in a window.
• cv2.imshow('edge', edge): Displays the edge-detected image in a window.

import cv2
img = cv2.imread("test.jpeg")
t_lower = 50  
t_upper = 150 
edge = cv2.Canny(img, t_lower, t_upper)
cv2.imshow('original', img)
cv2.imshow('edge', edge)
cv2.waitKey(0)
cv2.destroyAllWindows()

Output:

This is optional parameter that is used to specify the order of the Sobel filter used to calculate gradient in the Canny algorithm. We can increase Aperture size when we want to detect more detailed features.

• cv2.Canny(img, t_lower, t_upper, apertureSize=aperture_size): Applies Canny edge detection with a lower threshold (t_lower), upper threshold (t_upper) and Sobel filter aperture size (aperture_size) to detect edges in the image.
• aperture_size = 5: Defines the Sobel kernel size for gradient computation must be an odd number between 3 and 7 and a larger value captures more detailed edges.

import cv2
img = cv2.imread("test.jpeg") 
t_lower = 100  
t_upper = 200  
aperture_size = 5 
edge = cv2.Canny(img, t_lower, t_upper, 
                 apertureSize=aperture_size)
cv2.imshow('original', img)
cv2.imshow('edge', edge)
cv2.waitKey(0)
cv2.destroyAllWindows()

Output: 

It's a boolean parameter that specifies if we want to calculate usual gradient equation or the L2Gradient algorithm.

• aperture_size = 5: Specifies size of the Sobel kernel used to calculate the gradients and we set it to 5 here.
• L2Gradient = True: Enables more precise gradient calculation using the L2 norm.
• edge = cv2.Canny(img, t_lower, t_upper, L2gradient=L2Gradient): Applies the Canny edge detection algorithm to the image with the specified thresholds and L2 gradient

import cv2
img = cv2.imread("test.jpeg") 
t_lower = 100
t_upper = 200 
aperture_size = 5
L2Gradient = True 
edge = cv2.Canny(img, t_lower, t_upper, L2gradient = L2Gradient )
cv2.imshow('original', img)
cv2.imshow('edge', edge)
cv2.waitKey(0)
cv2.destroyAllWindows()

 Output: 

Here we will use both attributes within the same function.

import cv2 
img = cv2.imread("test.jpeg") 
t_lower = 100
t_upper = 200 
aperture_size = 5 
L2Gradient = True
edge = cv2.Canny(img, t_lower, t_upper,
                 apertureSize = aperture_size, 
                 L2gradient = L2Gradient ) 
cv2.imshow('original', img)
cv2.imshow('edge', edge)
cv2.waitKey(0)
cv2.destroyAllWindows()

Output: 

With the flexibility of the Canny() function we can experiment with different parameters to effectively detect edges and enhance the details in our images.

RetroSearch is an open source project built by @garambo | Open a GitHub Issue

Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo

HTML: 3.2 | Encoding: UTF-8 | Version: 0.7.4