Edge Detection Function Using Canny Operator with Comprehensive Code Implementation

This document provides a comprehensive overview of the edge detection function implementation using the Canny operator in MATLAB. This function serves as a method for detecting edges within digital images by employing the Canny operator, which represents a classical edge detection algorithm. The implementation follows a systematic approach: initially, the function applies Gaussian filtering to reduce image noise using MATLAB's imgaussfilt() function with configurable sigma parameters. Subsequently, it calculates gradient magnitude and direction for each pixel through Sobel operators implemented via imgradient() or manual convolution operations. The algorithm then performs non-maximum suppression to retain only local gradient maxima as potential edge points, involving neighborhood comparisons and interpolation for gradient directions. Finally, the implementation utilizes double thresholding with hysteresis processing to connect and filter edges, where strong edges are directly preserved while weak edges are retained only if connected to strong edges. This thresholding process typically employs edge() function with 'Canny' method or custom implementation using threshold values. Through detailed code annotations, users can better understand the function's implementation workflow, parameter significance (such as threshold ratios and sigma values), and algorithm optimization techniques, thereby enhancing comprehension and application capabilities of edge detection algorithms in practical scenarios.