A Robust Method for Handwriting Digital Recognition
Liu Yang · Journal of Wuhan Yejin University of Science and Technology · 2004
A new kind of sub-stroke pattern is proposed in this paper based on the two ends of the digital. A method of structural pattern matching is adopted to accomplish the off-line recognition of noise-containing digital. First,the number of the ends is calculated and the direction of the stroke is coded after the preprocessing of the digital picture. Then the digital is recognized by judging the intersections of horizontal(vertical) cutting and matching the sub-pattern, according to the number of the ends and the structural features of the digital. The results indicate that the method is robust to noise, easy to expand, and effective for off-line digital recognition .