New orthogonal multiwavelets offer improved image compression and denoising

Researchers have developed two novel orthogonal multiwavelets with supercompact support, utilizing the Fast Bauer's method for matrix spectral factorization. These new multiwavelets exhibit orthogonality and symmetry/antisymmetry, with one offering superior coding and smoothness compared to existing supercompact multiwavelets. Evaluations in edge detection, image compression, and signal denoising demonstrate that these new multiwavelets outperform other multifilters, including GHM, SA4, CL, Integer Haar, and Alpert, in metrics like SSIM and MS-SSIM. AI

IMPACT These advancements in multiwavelet filters could lead to more efficient and higher-quality image compression and signal denoising techniques.

RANK_REASON The cluster contains an academic paper detailing new methods and findings in computer vision. [lever_c_demoted from research: ic=1 ai=0.7]

Read on arXiv cs.CV →

• Alpert
• Fast Bauer's method
• Integer Haar
• MS-SSIM
• Structural Similarity Index Measure

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →