Compressed Sensing Image Compression with MATLAB Simulation

This research investigates image compression using compressed sensing technology with MATLAB simulations, aiming to explore a novel compression methodology that achieves higher compression ratios while retaining essential image information. The implementation involves applying compressed sensing principles to reduce image file sizes, conserve storage space, and minimize bandwidth consumption during transmission. The study thoroughly analyzes compressed sensing algorithm fundamentals—including sparse representation, measurement matrices, and reconstruction algorithms like L1-norm minimization—while evaluating performance through quantitative metrics such as PSNR and SSIM. Key MATLAB functions demonstrated include dwt2 for wavelet decomposition, randn for generating random measurement matrices, and l1magic toolbox functions for signal reconstruction. Experimental validation confirms the method's effectiveness and feasibility, contributing to advancements in image compression technology and providing reference value for related research fields.