Abstract: An online handwritten character recognizing system and method thereof, is provided which includes a feature extracting section for providing a Fourier transformation of the coordinate points of the character strokes that have been inputted through a tablet input/output section and normalized for distance and time by a preprocessing section. The Fourier coefficient is provided by converting a coordinate function expressing the time between inputted points into a frequency function. The recognizing system also includes a character recognizing section for dividing the Fourier coefficient provided by the feature extracting section into an imaginary part and a real part and multiplying a weight to the imaginary part and the real part and thereafter recognizing this as a character by utilizing perceptron, so that complicated Hangul (Korean alphabet) and English characters can be rapidly and precisely recognized.