计算机应用研究2009，Vol.26Issue(8)：2876-2878,2897,4.DOI:10.3969/j.issn.1001-3695.2009.08.021
基于小波和动态时间弯曲的主题查找算法
Finding motifs based on wavelet and dynamic time warping
摘要
Abstract
Since the proposed finding motifs algorithms could not combine the accuracy with complexity, and could not find the similar shapes, this paper proposed a finding motifs algorithm based on wavelet and dynamic time warping. Employed wavelet transform and a pruning strategy to reduce data dimension and complexity in motif finding, respectively. Because of its high accuracy, USED DTW to find the shape motifs. And combined with the formula v-shift, which ignored the vertical distance , this algorithm could find out the shapes which differed in size but actually similar to the predefined shape. Experiment shows that this algorithm not only has a higher matching accuracy and lower calculation cost, but also can find the shape motifs which are similar to the real world shape. It has great practical significance and there is no dismissal.关键词
形状数据库/主题/哈尔小波/动态时间弯曲Key words
database of shape/motif/Haar wavelet/DTW分类
信息技术与安全科学引用本文复制引用
宋晓宇,郭鲁,孙焕良..基于小波和动态时间弯曲的主题查找算法[J].计算机应用研究,2009,26(8):2876-2878,2897,4.基金项目
国家科技支撑计划资助项目(2008BAJ08B08) （2008BAJ08B08）
辽宁省自然科学(博士启动)基金资助项目(20071004) （博士启动）
辽宁省教育厅攻关计划资助项目 （）
