基于结构理解的笔迹智能编辑  

Intelligent Editing of Handwriting Based on Structure Understanding

作　　者：苏蕊[1] 张习文[2] 梁永全[3] 戴国忠[2] 

机构地区：[1]山东济宁医学院信息科学技术系,济宁272013 [2]中国科学院软件研究所人机交互技术与智能信息处理实验室,北京100080 [3]山东科技大学信息学院,青岛266510

出　　处：《系统仿真学报》2006年第z1期371-373,共3页Journal of System Simulation

基　　金：国家自然科学基金项目(60033020)。

摘　　要：随着人机界面朝着自然高效的方向发展,手写笔迹的结构分析和智能编辑越来越被人们所关注。给出了一种基于多层次相互作用的结构理解的笔迹智能编辑方法。该方法先自下而上初步提取单字、行、段,然后自上而下利用整体信息对组成信息进行精确分割,提取的信息表示成一个多层次的结构。基于笔迹的结构,用户可以使用笔手势对笔迹进行编辑。实验评估表明,该方法对笔迹的智能编辑具有较好的效果。Human-computer interface is developing along the direction of natural, intelligent and effective, the structure analysis and intelligent editing of handwriting is becoming more and more attractive. A new approach of intelligently editing handwriting was proposed based on multilayer structure understanding. The approach distills separate word, line, paragraph primarily from bottom to up, then makes use of holistic information to partition accurately the compositive information from top to bottom. The distilled information is expressed as a multilayer structure for the editing and recognizing of handwriting. The results of the informal user test show that the approach is effective and robust for intelligent editing of handwriting.

关 键 词：人机交互 笔式界面 模式识别 笔迹结构理解 智能编辑 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]