基于表象式语义网络的图匹配算法
首发时间:2009-01-08
摘要:提出了一种在表象式语义网络中查找的方法,表象式语义网络问题的求解一般都是通过图匹配实现的,首先根据待求解的问题的要求构造一个带变量节点的语义网络,然后与计算机视觉系统中己存储的语义网络进行图匹配。当语义网络中的询问部分与系统中的语义网络图匹配后,则与询问部分匹配的事实就是问题的解。图匹配问题可以通过构造一个图的附属数据结构来完成,这个附属数据结构也称为相连图(association graph),对于两个图G=(V,A)以及G’=(V’,A’),构造相联图G”=(V”,A”),也就是说,V”是所有可能节点匹配对的集合,A”是所有相容节点匹配的集合。这相当于在相联图中寻求一个最大的基团(clique),其中基团定义为G”的完全连通的一个子图。最大基团满足其节点集合不是任何其它基团节点集的适当子集。
For information in English, please click here
Algorithm of graph matching based on mental imagery semantic nets
Abstract:The basic thought of Partition-overall histogram is to divide the image space according to a certain strategy, and then calculate color histogram of each block as its color feature. Users choose the blocks that contain important space information, confirming the right value, and the system calculates the distance between the corresponding blocks, Other blocks merge into part overall histograms and calculate the distance again. Then accumulate all the distance as the real distance between two pictures. The less distance there is, the less visional difference they have. Experiment results show that this algorithm can obtain more optimized system with higher Precision and Recall and better than traditional histogram and Partition-based histogram.
Keywords: Artificial intelligence mental imagery semantic nets graph matching
论文图表:
引用
No.2748938438712314****
同行评议
勘误表
基于表象式语义网络的图匹配算法
评论
全部评论0/1000