🖼️ SVD Image Compression

Compress images using Singular Value Decomposition

Number of Singular Values: 10

Original Image

Compressed Image (Rank 10)

Compression Ratio

-

Retained Energy

-

PSNR (dB)

-

Singular Values Decay

How SVD Compression Works

Singular Value Decomposition (SVD) decomposes an image matrix A into three matrices: A = UΣV^T

U: Left singular vectors (image space basis)
Σ: Diagonal matrix of singular values (importance weights)
V: Right singular vectors (feature space basis)

By keeping only the k largest singular values, we can approximate the image with much less data while preserving the most important features.