opencvEdge detection


  • edges = cv2.Canny(image, threshold1, threshold2[, edges[, apertureSize[, L2gradient]]])
  • void Canny(InputArray image, OutputArray edges, double threshold1, double threshold2, int apertureSize=3, bool L2gradient=false


imageInput image
edgesOutput image
threshold1First threshold for hysteresis procedure
threshold2Second threshold for hysteresis procedure
apertureSizeAperture size for Sobel operator
L2gradientFlag indicating whether a more accurate algorithm for image gradient should be used

Canny algorithm

The Canny algorithm is a more recent edge detector designed as a signal processing problem. In OpenCV, it outputs a binary image marking the detected edges.


import cv2
import sys

# Load the image file
image = cv2.imread('image.png')

# Check if image was loaded improperly and exit if so
if image is None:
    sys.exit('Failed to load image')

# Detect edges in the image. The parameters control the thresholds
edges = cv2.Canny(image, 100, 2500, apertureSize=5)

# Display the output in a window
cv2.imshow('output', edges)

Canny Algorithm - C++

Below is an usage of canny algorithm in c++. Note that the image is first converted to grayscale image, then Gaussian filter is used to reduce the noise in the image. Then Canny algorithm is used for edge detection.

// CannyTutorial.cpp : Defines the entry point for the console application. 
// Environment: Visual studio 2015, Windows 10
// Assumptions: Opecv is installed configured in the visual studio project
// Opencv version: OpenCV 3.1

#include "stdafx.h"

int main()

    //Modified from source:
    cv::Mat imgOriginal;        // input image
    cv::Mat imgGrayscale;        // grayscale of input image
    cv::Mat imgBlurred;            // intermediate blured image
    cv::Mat imgCanny;            // Canny edge image

    std::cout << "Please enter an image filename : ";
    std::string img_addr;
    std::cin >> img_addr;

    std::cout << "Searching for " + img_addr << std::endl;

    imgOriginal = cv::imread(img_addr);            // open image

    if (imgOriginal.empty()) {                                    // if unable to open image
        std::cout << "error: image not read from file\n\n";        // show error message on command line
        return(0);                                                // and exit program

    cv::cvtColor(imgOriginal, imgGrayscale, CV_BGR2GRAY);        // convert to grayscale

    cv::GaussianBlur(imgGrayscale,            // input image
        imgBlurred,                            // output image
        cv::Size(5, 5),                        // smoothing window width and height in pixels
        1.5);                                // sigma value, determines how much the image will be blurred

    cv::Canny(imgBlurred,            // input image
        imgCanny,                    // output image
        100,                        // low threshold
        200);                        // high threshold

    // Declare windows
    // Note: you can use CV_WINDOW_NORMAL which allows resizing the window
    // or CV_WINDOW_AUTOSIZE for a fixed size window matching the resolution of the image
    // CV_WINDOW_AUTOSIZE is the default
    cv::namedWindow("imgOriginal", CV_WINDOW_AUTOSIZE);        
    cv::namedWindow("imgCanny", CV_WINDOW_AUTOSIZE);

    //Show windows
    cv::imshow("imgOriginal", imgOriginal);        
    cv::imshow("imgCanny", imgCanny);

    cv::waitKey(0);                    // hold windows open until user presses a key
    return 0;

Calculating Canny Thresholds

Automatic calculation of low and high thresholds for the Canny operation in opencv

Canny Edge Video from Webcam Capture - Python

import cv2

def canny_webcam():
    "Live capture frames from webcam and show the canny edge image of the captured frames."

    cap = cv2.VideoCapture(0)

    while True:
        ret, frame =  # ret gets a boolean value. True if reading is successful (I think). frame is an
        # uint8 numpy.ndarray

        frame = cv2.GaussianBlur(frame, (7, 7), 1.41)
        frame = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)

        edge = cv2.Canny(frame, 25, 75)

        cv2.imshow('Canny Edge', edge)

        if cv2.waitKey(20) == ord('q'):  # Introduce 20 milisecond delay. press q to exit.


Canny Edge Thresholds prototyping using Trackbars

CannyTrackbar function allows for a better understanding of 
the mechanisms behind Canny Edge detection algorithm and rapid
prototyping. The example includes basic use case.

2 of the trackbars allow for tuning of the Canny function and
the other 2 help with understanding how basic filtering affects it.
import cv2

def empty_function(*args):

def CannyTrackbar(img):
    win_name = "CannyTrackbars"

    cv2.resizeWindow(win_name, 500,100)

    cv2.createTrackbar("canny_th1", win_name, 0, 255, empty_function)
    cv2.createTrackbar("canny_th2", win_name, 0, 255, empty_function)
    cv2.createTrackbar("blur_size", win_name, 0, 255, empty_function)
    cv2.createTrackbar("blur_amp", win_name, 0, 255, empty_function)

    while True:
        cth1_pos = cv2.getTrackbarPos("canny_th1", win_name)
        cth2_pos = cv2.getTrackbarPos("canny_th2", win_name)
        bsize_pos = cv2.getTrackbarPos("blur_size", win_name)
        bamp_pos = cv2.getTrackbarPos("blur_amp", win_name)

        img_blurred = cv2.GaussianBlur(img.copy(), (trackbar_pos3 * 2 + 1, trackbar_pos3 * 2 + 1), bamp_pos)
        canny = cv2.Canny(img_blurred, cth1_pos, cth2_pos)
        cv2.imshow(win_name, canny)

        key = cv2.waitKey(1) & 0xFF
        if key == ord("c"):

    return canny

img = cv2.imread("image.jpg")
canny = CannyTrackbar(img)
cv2.imwrite("result.jpg", canny)