| 基于相关反馈技术的在线复杂图形识别 |
| |
| 引用本文： | 王强,孙正兴.基于相关反馈技术的在线复杂图形识别[J].东南大学学报,2005,21(2):153-158. |
| |
| 作者姓名： | 王强 孙正兴 |
| |
| 作者单位： | 南京大学计算机软件新技术国家重点实验室，南京210093 |
| |
| 基金项目： | 国家自然科学基金，Program for New Century Excellent Talents in University |
| |
| 摘 要： | 提出一种在线复杂图形识别方法,在该方法中引入相关反馈技术来逐渐捕获用户绘图意图并使用用户模型来适应不同用户的绘图习惯.首先,在根据图形向量化特征的相似度计算给出候选识别结果的基础上,利用相关反馈技术不断降低模糊性而提高识别效果.其次,通过记录和分析"历史信息"动态的为不同用户建立用户模型,从而适应不同的用户习惯和绘图意图.同时,引入了一个基于模型的动态匹配策略.实验证明所介绍的识别方法在试验中取得了很好的效果.
|
| 关 键 词： | 基于草图的用户接口 在线复杂图形识别 动态用户模型 相关反馈 |
Online composite shape recognition based on relevance feedback |
| |
| Wang Qiang,Sun Zhengxing.Online composite shape recognition based on relevance feedback[J].Journal of Southeast University(English Edition),2005,21(2):153-158. |
| |
| Authors: | Wang Qiang Sun Zhengxing |
| |
| Abstract: | This paper describes a novel method of online composite shape recognition in terms of the relevance feedback technology to capture a user's intentions incrementally,and a dynamic user modeling method to adapt to various users' styles.First,the relevance feedback is adapted to refine the recognition results and reduce the ambiguity incrementally based on the establishment of a feature-based vector model of a user's sketches.Secondly,a dynamic user modeling is introduced to model the user's sketching habits based on recording and analyzing historical information incrementally.A model-based matching strategy is also employed in the method to recognize sketches dynamically.Experiments prove that the proposed method is both effective and efficient. |
| |
| Keywords: | sketchy-based user interface online composite shape recognition dynamic user modeling relevance feedback |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
