Many　practical　applications　need　to　eliminate　noise　from　noisy　images.　This　paper　first　introduce　a　simplified　support　vector　regression　(SSVR)　method　for　single　image　denoising　under　mixture　noise　environments.　The　proposed　SSVR　method　has　a　lower　computational　complexity　than　related　SVR　approaches　to　image　denoising.　Furthermore,　we　propose　a　novel　multiple　SVR　approach　for　multi-image　denoising　by　extending　the　SSVR　method.　Experiment　results　show　that　the　proposed　SVR-based　approach　has　good　performance　in　fast　removing　mixture　noise.　©　2011　IEEE.