Driver　fatigue　detection　is　an　important　research　content　in　the　field　of　HCI　(Human-Computer　Interaction)　and　CV　(Computer　Vision).　For　the　effectiveness　about　the　face　image　expression　using　Gabor　feature,　many　researchers　introduced　the　Gabor　wavelet　to　the　recognition　of　fatigue.　The　dimension　of　Gabor　feature,　however,　is　too　high,　besides　the　computation　and　memory　requirements　are　too　large,　so　it　needs　dimensional　reduction.　In　this　paper,　we　present　a　dynamic　Gabor　feature　fatigue　analysis　method　based　on　features　fusion.　First,　Gabor　filters　are　employed　to　extract　multi-scale　and　multi-orientation　features　from　each　image　in　the　face　image　sequences.　Then　we　threshold　Gabor　features　and　fused　them　based　the　encoding　manner.　Finally,　dynamic　face　features　that　after　fused　are　extracted　for　fatigue　detection.　The　method　in　this　paper　is　able　to　reduce　the　dimension　of　features　effectively,　and　also　acquire　the　satisfactory　experimental　results.　©　2010　IEEE.