This　paper　proposed　a　fast,　algorithm　for　2-D　ARMA　parameter　estimation　under　noise　environments.　The　proposed　algorithm　is　based　on　a　kind　of　inverse　filter　technique　for　the　approximation　of　random　excitation.　The　proposed　algorithm　only　solves　a　low　order　2-D　AR　system　for　the　2-D　ARMA　parameter.　In　contrast　with　Zhang　and　Cheng　＇s　and　Kizilkaya　and　Kayran＇s　algorithms　for　2-D　ARMA　parameter　estimation,　the　proposed　algorithm　considers　the　noisy　observation　and　is　more　suitable　for　practical　applications.　Moreover,　the　proposed　algorithm　has　a　fast;　speed　due　to　a　low　computational　cost.　Simulation　results　show　that,　the　proposed　algorithm　can　obtain　more　accurate　estimates　with　a　faster　speed　than　the　two　algorithms.