由手写体图象恢复笔划的一种方法

Approach to strokes recovering from handwriting images

本文提出一种由手写体图象复恢笔划是的跟踪算法。该算法突破了象素级链码跟踪的传统方法，在考虑了周围笔划的走向基础上进行笔划跟踪。首先，灰度图象经过二值化和细化产生手写体的原始骨架；其次由手写体的宽度信息进一步得到两类骨架：基本骨架与参考骨架；再通过基本骨架的跟踪连接得到笔划段。实验表明该方法对手写体笔划段。最后讨论了跟踪后处理，其目的是为了得到和人书写结果类似的笔划。实验表明该方法对手写体笔划跟踪有很好的效果。

his paper presents a novel stroke tracing approach to recovering the strokes from the handwriting images. It is carried out not only based on pixel by pixel, but considering the local relation of handwriting strokes. Firstly, the primary skeletons of handwriting are produced through the binarization and thinning. Secondly, after the basic skeletons and reference skeletons are identified based on the width information of handwriting, the strokes segments are obtained by connecting and tracing on the basic skeletons. Finally, for making the recovered strokes more natural, the postprocessing of tracing is discussed. Experimental results show the effectiveness of the approach.