摘要
研究了基于特征点的图形检索方法。忽略通常意义下的图元及图元之间的相互关系,采用特征点对图形进行表述。以基准点和基准向量为出发点,用特征点与基准点之间的距离信息,特征向量与基准向量的角度信息将图形中的各个特征点唯一确定,然后分析距离信息和角度信息。提出一种寻找最接近特征点的方法,并以此为基础对数据进行整合计算,计算出图形的数字特征,完成检索。实验证明了该方法的有效性。
This paper deals with the graphics searching method based on limited feature points. Instead of focusing on the relationship among the graphic elements, the feature points are selected to describe the graphics. All the feature points are uniquely determined by using of the distance between the feature point and datum point and the angle between the datum vector and the feature vector. Based on the analysis of those data, a method is provided for searching the most approximate feature points. The data are computed and integrated, and the numerical characteristic of the graphics is searched. Experiments show that the graphics searching method is valid.
出处
《图学学报》
CSCD
北大核心
2014年第1期16-20,共5页
Journal of Graphics
关键词
图形检索
近似特征点
数字特征
特征向量
searching graphic
approximate feature point
numerical characteristic
feature vector
