Scale　invariant　feature　transform　(SIFT)　is　a　local　feature　descriptor　that　searches　from　extreme　points　in　spatial　scales.　Aiming　at　the　speeding　up　the　matching　process　of　the　traditional　Scale-invariant　Feature　Transform　(SIFT)　algorithm,　a　fast　image　matching　method　based　on　improved　SIFT　is　proposed.　It　eliminates　some　invalid　feature　points　by　calculating　the　two-dimensional　entropy　of　feature　points.　in　the　feature　extraction　stage,　eliminating　the　feature　points　with　larger　two-dimensional　entropy.　In　the　feature　point　matching　stage,　a　reliable　matching　pair　is　found　by　traversing　the　difference　of　two-dimensional　entropy　between　a　feature　point　of　the　reference　image　and　a　feature　point　of　the　image　to　be　matched.　this　will　reduce　the　computation　of　Euclidean　distance　in　matching　process.　The　experimental　results　show　that　the　improved　algorithm　has　a　capacity　to　effectively　improve　the　image　matching　speed　on　the　premise　of　maintaining　stability,　and　it　is　1.6　times　as　much　as　the　original　matching　speed.　©　2019,　The　Press　of　NUC.　All　right　reserved.