In this research, we proposed a method for the stroke extraction of static Chinese character. A Chinese character could be separated into several parts. In order to recognize every independent part, we adopted the binary image of Chinese character and connected component labeling methods. First, the strokes of connection were marked after categorizing to the same area in order. The boundary of every independent area that was also be called the contour of part will be identified and the pixels of the contour were assigned numbers in order. Second, the breaking point of junction of stroke was identified dilation method. Drawing the split line after the braking points were matched. Finally, we could successfully extract the stroke
