Holographic Image Compression via Sparse Representation and Orthogonal Pursuit in Transformed Spaces

This paper introduces an innovative image compression framework based on Compressed Sensing theory and sparserepresentation. The proposed algorithm utilizes an iterativedecomposition via Orthogonal Matching Pursuit (OMP) appliedto an overcomplete dictionary generated through Kroneckerproducts of DCT bases. The system is highly configurable,allowing for a dynamic trade-off between fidelity and size.Compared to standard formats, the method achieves a baselinecompression ratio of 28:1 with a PSNR of 35.85 dB, whilemaintaining the capability to scale up to 150:1 for ultra-lowbandwidth scenarios.