Based on the analysis of above drawbacks , this paper proposes frequent access pattern tree algorithm ( fapt ) . this algorithm includes two steps : access pattern tree method , through the pattern matching method it saves user ' s visit sequences with tree ; pruning method , it uses frequent degree to prune access pattern tree which is under the frequent degree 在分析以上不足的基础上,提出了频繁访问模式树( fapt )算法,该算法包括以下两个步骤:访问模式树的生成,通过模式匹配的方法将用户访问序列以树形结构来存储;修剪的策略,利用频繁度对访问模式树进行修剪,修剪掉其中低于频繁度的节点。
In feature matching , a dot pattern matching method is developed , which combine subschema matching method for solving rotations and scale changes problems , and employ variable bounding box to solve translations problem . then the problems of arbitrary translations , rotations , and scale changes in feature matching have been well 特征匹配算法采用了基于点模式的匹配方法,并结合基于结构的匹配方法解决图像的旋转、平移问题,同时采用可变限界盒方法解决图像的变形问题,使得指纹图像匹配中因图像旋转、平移、形变等带来的问题得到了较好的解决。