Air-Landed Airplane Recognition Based on EHMM

Proposes an automatic strategy to estimate the hidden states number and embedded hidden states number of EHMM,so as to estimate the optimal model structure.The algorithm is ap-plied to air-landed airplane recognition,after comparing the EHMM model structure with other EHMM models,it can conclude,that the EHMM model structure achieves higher recognition rate(almost to 100%) than other's,and costs fewer system resource and com-puting time(almost shorten 30%).