Age　estimation　is　an　important　enabling　capability　for　its　potential　applications　in　the　near　future,　especially　in　supervision　of　minors,　retrieving　photograpgs　at　specific　age　ranges　in　e-photoalbums,　and　predicting　future　facial　appearance　of　subjects.　Due　to　the　intrinsic　and　external　factors　the　age　estimation　is　a　very　complex　task.　One　of　the　most　important　steps　is　feature　extraction.　However,　Single　feature　extraction　methods　are　commonly　used,　and　all　of　them　have　their　own　drawbacks　and　limitations.　In　this　paper,　LBP　combined　with　sp-PCA[1]　method　for　age　feature　extraction　is　proposed.　Addtionally　in　order　to　make　the　intra-class　more　compact,　extra-class　more　loose,　We　adopt　the　Relevant　Component　Analysis(RCA)　for　distance　metric　adjustment.