Visual　saliency　reflects　that　we　always　focus　on　the　significant　region　of　a　scene　and　processes　it　preferentially.　In　this　paper　we　present　a　3D　face　recognition　algorithm　based　on　the　analysis　of　significant　features.　It　separated　the　target　region　and　surrounding　environment　of　human　face　and　highlighted　the　target　region　evenly.　After　the　fusion　of　texture,　depth　and　normal　features,　we　got　91.3%　accuracy.　Copyright　©　2014　Binary　Information　Press.