This paper considers the problem of H-infinity model reduction problem with finite frequency (FF) ranges of the input vector for two-dimensional (2D) continuous systems. Given an asymptotically stable system, the main objective is to find a stable reduced-order model such that the error of the transfer functions between the original system and the reduced-order one is bounded over a FF range. Using the well known generalized Kalman Yakubovich Popov (gKYP) Lemma and the Finsler's Lemma, sufficient conditions for the existence of H-infinity model reduction for different FF ranges are proposed and then unified in terms of solving a set of linear matrix inequalities (LMIs). Finally, the effectiveness of the proposed method is illustrated by a numerical example.