Extracting Entire Image Edge Pixels Using Optimal Canny Edge Detection Operator

This MATLAB program implements the extraction of complete image edge pixels using the Canny operator, recognized as optimal for edge detection. The process consists of three main computational stages. First, Gaussian filtering is applied to eliminate noise and improve the signal-to-noise ratio, typically implemented using MATLAB's fspecial('gaussian') function with appropriate standard deviation parameters. Second, first-derivative Gaussian filtering detects image edges by calculating gradients using Sobel or Prewitt operators through the imgradient function. These two operations are efficiently combined into the unified Canny operator workflow. Finally, morphological operators such as dilation and closing (implemented via imdilate and imclose functions) connect discontinuous edges from thresholded images to form complete edge contours. The implementation may include pre-processing steps like contrast adjustment using imadjust and brightness normalization through histogram equalization (histeq function) to enhance detection accuracy. The algorithm effectively handles edge thinning and hysteresis thresholding to distinguish strong and weak edges while minimizing false detections.