To　solve　the　problems　of　low　spatial　resolution　and　much　mixed　pixel　of　traditional　hyperspectral　image,　this　paper　puts　forward　a　hyperspectral　image　fusion　method　based　on　Gram-Schmidt　improved　algorithm　(GS3),　in　order　to　improve　the　fusion　effect　of　the　hyperspectral　remote　sensing　image　and　high　spatial　resolution　image.　It　adopts　GS3　and　other　various　fusion　methods　to　make　fusion　comparison　experiment　on　the　Hyperion　hyperspectral　image　and　ALI　panchromatic　band　of　the　same　time　phase　and　the　same　sensor　(EO-1)　and　respectively　uses　the　qualitative　and　quantitative　methods　to　make　comprehensive　analysis　and　evaluation　of　the　fusion　results.　The　experimental　results　show　that,　compared　with　other　hyperspectral　image　fusion　methods,　GS3　method　can　better　maintain　the　spectral　feature　and　space　texture　feature　of　hyperspectral　image　at　the　same　time,　which　is　a　relatively　ideal　hyperspectral　image　fusion　method.　©　2013　Asian　Network　for　Scientific　Information.