节点文献
一种自适应连续手绘线语义识别算法
A Self-Adaptive Algorithm on Continuous Sketching Recognition
【摘要】 针对手绘交互领域连续手绘线的语义识别问题提出了一种适用性广,调试方便且对不同软件平台具有自适应性的识别算法:利用手绘线的整体输入信息,在保留所有原始采样点的前提下,对手绘线重采样后再进行分段得到特征字符串,避免了平台相关性。之后,根据输入方向拓展了基本动作库,并建立所需要的动作库。利用正则表达式工具,将特征字符串与动作库中的模式串进行匹配翻译。通过实验,文章还讨论了该算法的优缺点及适用性。
【Abstract】 A self-adaptive algorithm for different software platforms on continuous sketching recognition is presented.Depending on the data of whole stroke,it is resampled to avoid platform-related problem.A basic gesture library is extended by input direction and the aimed gesture library is built.The matching work is done using regular expression grammar.Comparing the results in two different platforms,applicability is also discussed.
【关键词】 计算机应用;
手绘识别;
语义识别;
自适应性;
【Key words】 computer application; sketch recognition; sketch understanding; self-adaptability;
【Key words】 computer application; sketch recognition; sketch understanding; self-adaptability;
【基金】 中国传媒大学研究生教育改革工程资助项目(CX20080507)
- 【文献出处】 工程图学学报 ,Journal of Engineering Graphics , 编辑部邮箱 ,2010年05期
- 【分类号】TP391.41
- 【被引频次】1
- 【下载频次】95