Diabetic　retinopathy　contributes　to　serious　health　problem　in　many　parts　of　the　world.　With　the　motivation　of　the　needs　of　the　medical　community　system　for　early　screening　of　diabetics　and　other　diseases　a　computer　aided　diagnosis　system　is　proposed.　This　work　is　aimed　to　develop　an　automated　system　to　analyze　the　retinal　images　for　important　features　of　diabetic　retinopathy　using　image　processing　techniques　and　an　image　classifier　based　on　BP　neural　network　which　classify　the　images　according　to　the　disease　conditions.　In　the　course　of　the　study,　based　on　the　actual　problem　and　BP　neural　network　characteristics,　we　exploit　an　improved　BP　neural　network　classification　method.