Estimating the number of components by a data independent penalty function approach

We consider the estimation procedure of the number of components of a two-dimensional sinusoidal model which has wide applications in statistical signal processing and in texture classifications. The proposed method is based on the penalty function approach like other information theoretic criteria. The new method is shown to be consistent and we provide the performance analysis of the proposed method. It is observed that the performance analysis can be used to provide a data independent penalty function from a class of penalty functions. The effectiveness of the proposed method is verified using computer simulations.