Comparative Analysis of Six Edge Detection Operators (Gabor, Laplacian, Prewitt, Robert, Sobel, Wallis) on Three Image Types

In this documentation, I present a crucial program from my graduation project that implements and compares six distinct edge detection operators (Gabor, Laplacian, Prewitt, Robert, Sobel, Wallis) applied to three different image types. The implementation features optimized convolution kernel designs for each operator: Gabor filters with multi-orientation frequency tuning, Laplacian of Gaussian (LoG) for second-order derivative detection, and efficient 3x3 kernel implementations for Prewitt, Robert, and Sobel operators. The Wallis filter incorporates local statistical normalization for enhanced edge preservation. Through extensive research and practical experimentation, this program demonstrates robust edge detection capabilities with clear, accurate results that were fundamental to my graduation research. The code architecture allows for systematic parameter optimization and performance benchmarking across different image characteristics. I'm sharing this implementation hoping it will prove valuable for your research projects or practical applications in computer vision and image processing.