Image Feature Extraction Using Fourier Transform

In this article, we explore how to perform spectral analysis of signals using Fourier transform. First, we demonstrate the Fourier spectrum diagram using MATLAB's fft2 function for 2D Fourier transform and fftshift to center the frequency components, providing better visualization of frequency domain characteristics. Next, we introduce frequency domain low-pass filtering methods, implementing ideal low-pass filters with specified cutoff frequencies, Gaussian low-pass filters with sigma parameters controlling blur intensity, and Butterworth filters with order parameters affecting transition sharpness. We compare processing results under different filter functions and parameters to analyze their effects on image smoothing and noise reduction. We also investigate frequency domain high-pass filtering techniques, including ideal high-pass filters for edge detection, Gaussian high-pass filters for gradual frequency attenuation, and Butterworth high-pass filters for customizable frequency response. The comparative analysis examines how different filter functions and parameters enhance image details and edges. Through these experiments implemented using MATLAB's image processing toolbox, we gain deeper understanding of frequency domain filtering principles and applications, establishing a solid foundation for future research and practical implementations in digital image processing.