Infrared/millimeter　wave　(IR/MMW)　composite　guidance　is　currently　a　hot　research　topic　in　the　field　of　multi-mode　composite　guidance　technology.　The　penetrability　of　infrared　detection　system　is　less　than　that　of　passive　millimeter　wave　under　low　visibility　condition,　but　the　image　resolution　of　the　passive　millimeter　wave　(PMMW)　detection　is　better　than　that　of　infrared　images.　In　order　to　better　identify　the　target　contour　information,　in　this　paper,　a　new　edge　feature　fusion　method　was　put　forward　based　on　wavelet　packet　analysis　and　principal　component　analysis.　The　new　method,　firstly,　uses　wavelet　packet　edge　detection　to　get　three　edge　feature　images　which　contain　horizontal,　vertical　and　diagonal　edge　feature　respectively.　Secondly,　the　edge　feature　images　are　de-noised　using　wavelet　packet　de-noising　algorithm.　Thirdly,　the　edge　feature　images　are　fused　with　principal　component　analysis　method.　Lastly,　it　is　to　get　the　fusion　edge　feature　image　with　threshold　method.　The　experimental　results　indicate　that　compared　with　traditional　edge　detections　of　wavelet　analysis　and　wavelet　packet　analysis,　the　new　fusion　method　can　better　recognize　the　contour　information　of　the　target　object.　©,　2015,　China　National　Publishing　Industry　Trading　Corporation.　All　right　reserved.