Face　image　super-resolution　reconstruction　(SRR)　can　be　widely　used　in　forensic　analysis　and　video　surveillance.　With　the　recognition-based　idea,　the　statistical　characteristics　of　face　images　are　investigated　and　incorporated　into　SRR.　Based　on　the　set　theoretic　formulation,　a　projection　onto　convex　sets　(POCS)　algorithm　is　applied　to　find　the　solution　to　face　image　reconstruction.　Compared　with　the　traditional　POCS　based　SRR　methods,　the　proposed　approach　imposes　some　extra　constraint　sets　to　the　solution.　The　experiment　results　on　frontal　face　images　show　that　the　proposed　approach　gains　a　better　performance　both　on　noise　suppression　and　reconstruction　quality　and　has　the　advantage　of　simplicity　in　computation.