Pulse　Coupled　Neural　Network　(PCNN)　and　its　modified　forms　are　widely　applied　to　image　segmentation.　However,　most　PCNN　based　methods　are　difficult　in　segmentation　of　the　regions　that　contain　rich　texture　information　in　a　complex　image.　This　paper　proposes　an　improved　approach　to　image　segmentation　based　on　simplified　region　growing　PCNN　with　adaptive　texture　features　input.　It　shows　promising　results　to　process　regions　with　complex　texture.