In this paper, we analyzed the relationship of interfaces of DirectShow and show it using UML. Then we introduce the step how to implement capture images. ( 2 ) Composing handwriting font and recording some position. 文中详细分析了DirectShow常用接口之间的关系,并给出了UML类图,并在此基础上介绍了图像采集的步骤。(2)合成手写字体并记录书写过程中的相关信息。
Experimental results show that the proposed algorithm performs well, even for captions with unknown font, size, and position. 实验结果表明,即使在复杂背景下,对于字体、大小和位置都不确定的字幕,该算法仍具有良好的分割效果。