Target　recognition　in　remote　sensing　image　is　a　complicated　pattern　recognition　task.　In　this　paper　we　propose　a　new　method　of　recognition　of　bridge　over　water　in　remote　sensing　image.　Various　techniques　were　applied　to　the　process.　Image　segmentation　and　mathematics　morphology　techniques　were　used　to　find　out　water　areas　and　locate　the　ROIs(Regions　of　Interest),　and　then　feature　extraction　and　HNN　were　utilized　to　recognize　the　bridge.　During　the　process,　we　also　introduced　a　workable　method　of　binary　coding　specific　to　Discrete　Hopfield　Neural　Network.　The　results　show　that　the　process　we　propose　is　an　effective　method　to　fulfil　the　task.　The　main　contribution　of　this　paper　lies　in　successfully　applying　Discrete　Hopfield　Neural　Network,　which　few　scholars　used　before,　in　recognition　of　bridge　in　remote　sensing　image.　©　2011　IEEE.