As　a　key　technique　in　network　multimedia　signal　processing,　video　transcod-　ing　becomes　a　hot　topic　in　recent　years.　This　paper　presents　a　fast　intra　mode　deci-　sion　scheme　for　down-sizing　video　transcoding　in　H.264　based　on　hybrid　characteristic　of　multi-scale　videos.　In　order　to　reduce　the　high　computational　complexity　of　using　con-　ventional　intra　prediction　in　the　H.264　re-encoder,　the　proposed　scheme　firstly　utilizes　2D-histogram　to　extract　the　spatial　characteristic　of　macro-blocks　in　the　downsized　video　to　choose　from　intra　16×16　and　intra　4×4.　Then　Support　Vector　Machine　(SVM)　is　used　to　exploit　the　correlation　between　coding　information　extracted　from　the　input　high-　resolution　bit-stream　and　the　coding　modes　of　macro-blocks　in　down-sized　video　frames.　After　the　SVM　classifier,　improbable　modes　in　the　nine　intra　4×4　modes　are　eliminated　and　only　a　small　number　of　candidate　modes　are　carried　out　using　the　RDO　operations.　Hence,　remarkable　computing　time　can　be　saved,　up　to　74%,　while　maintaining　nearly　the　same　quality　of　the　transcoded　pictures.　©　2012　ISSN　2073-4212.