Abstract:Signal can be decomposed sparsely and power-focally in an over-complete dictionary with Matching Pursuit (MP). This paper proposes a modified MP method to decompose signal more sparsely. In the iteration procedure of the modified MP, the over-complete dictionary is classified into two separate dictionaries with the selected and unselected atoms, the algorithm is designed to have more chances than the original MP to choose the atom in the selected atom dictionary as the optimal atom by a simulate annealing threshold function, thus the algorithm availed for a more sparse decomposition. The decomposition results for a cosine-modulated exponential signal and an actual speech signal show that the proposed modified MP can decompose signal more sparsely.