an　inte grated　four－component　model－based　decomposition 　of polarimetric　sar　with　covariance　matrix

in 　this　paper we　developed　a　four－component　model－based 　incoherent 　decomposition　withcovariance　matrix　inte grating　selective 　de－orientation　and　generalized　volume　scattering．firstlyanew　cross －polarization 　coefficient 　is 　proposed　which　can　be　used 　for 　suppressing　the　helix　scatteringand　de－orientating　the　non－reflection 　symmetric　tar gets．secondl yyamaguchi 　decomposition 　is 　im－proved 　through　adoptin g　ageneralized　volume　scattering　model　instead 　of 　the　ori ginal 　one．the　newvolume　model　can　vary　adaptivel y　with　the　power　ratio　between　hh　and　vv．lastlythe　power　constrain 　is 　utilized 　to 　eliminate 　the　negative 　power　completel y．the　effectiveness 　of 　four－component　decomposition 　is 　demonstrated　lband　e－sar　and　airsar　polarimetric 　data．the 　results 　suggest 　thatcannot 　onl y　eliminate 　the　negative 　phenomenon　completel ybut　also 　can　enhance　double－bounce　scattering　in　urban　areas 　remarkabl y．it　can　effectivel y　suppress 　the　helix　components　which　are 　difficultto　produce 　in　most　natural 　areas