MATLAB Source Code for Canny Edge Detector

The following code presents a MATLAB implementation of edge detection using the Canny operator. Developed by John F. Canny in 1986, the Canny edge detector is a multi-stage algorithm widely recognized for its ability to identify image edges while effectively minimizing noise interference. The algorithm's key stages include Gaussian filtering, gradient computation, non-maximum suppression, and double thresholding. Our implementation leverages MATLAB's built-in functions to simplify the algorithm's execution while maintaining computational efficiency. For deeper understanding of Canny's theoretical foundations, users are encouraged to consult relevant literature and modify the code for experimental purposes.

```matlab

% Read input image using imread function

img = imread('lena.png');

% Convert RGB image to grayscale using rgb2gray

gray_img = rgb2gray(img);

% Apply Gaussian filter with standard deviation 1 for noise reduction

gaussian_img = imgaussfilt(gray_img, 1);

% Compute gradient components using gradient function

[Gx,Gy] = gradient(gaussian_img);

% Calculate gradient magnitude through Euclidean norm

G = sqrt(Gx.*Gx + Gy.*Gy);

% Perform non-maximum suppression with 1.5 pixel radius

nms_img = nonmaxsup(G, atan2(Gy, Gx), 1.5);

% Set double thresholds: high threshold at 10% of maximum intensity

high_thresh = max(nms_img(:)) * 0.1;

% Low threshold as 5% of high threshold for hysteresis thresholding

low_thresh = high_thresh * 0.05;

% Apply hysteresis thresholding to finalize edge detection

edge_img = hysthresh(nms_img, high_thresh, low_thresh);

% Display resulting edge map using imshow function

imshow(edge_img);

```