Provided is an improved apparatus and method for recognizing pattern data. The method including extracting a high frequency component with Y data from pattern data sensed through a camera equipped in a mobile station to more clearly recognize edge portions. The high frequency component and the Y data are weighted with predetermined weights, and input data is generated using the high frequency component and Y data weighted with the pre-set weights. Accordingly, edge portions of the input data are more clearly defined, thereby increasing a recognition rate of the pattern data.