Estimating of geometrical theory of diffraction (GTD) model parameters accurately and efficiently is significant for automatic target recognition.In this paper

a time-domain sparse component analysis method is designed for parameter estimation of GTD model.Firstly

potential locations of scattering centers

as well as columns of the dictionary

are reduced according to the high resolution range profile.Secondly

the time domain responses of GTD model is used to construct the dictionary

and the small amplitude areas are cut off.Compared with the presented method

not only columns of the new dictionary are reduced

but also the dictionary becomes a sparse matrix.As a consequence

it needs less memory to store

which is helpful to decrease computational complexity of component analysis.In addition

a reconstruction method based on the orthogonal matching pursuit is designed for our dictionary.Finally

experimental results demonstrate the performance and efficiency of the proposed method.