摘要: 图像的基于分块重量索引表示方法与基于重心位置索引表示方法具有相同的树状 结构,把这两种结构结合起来,构造出图象的特征向量(该特征向量即为高维空间中的点) ,为每个特征向量定义两种长度,构造图像的特征向量(即高维空间点集)的最小平面包围 盒,完成一种简单高效的图像索引查找方案。这种索引方法对污染、破损、变形等图像具有 较好的查找效果,同时也是在图像点集上定义多种距离的尝试。
关键词:
图像数据库,
R-树,
图像重量与重心
Abstract: A simple and efficient image index method based on weight and center of image gravity is proposed in the paper. Divided the image into small section s layer by layer, contructed the weight and the center sequence of the image (no ted IGAC), two kinds of distance of IGAC, and the concept of search box is given . Lot of experiment proves the method be efficient. At the end of the paper, mul ti-distance research problem is discussed.
中图分类号:
于万波,赵晶,李淑敏,魏小鹏. 基于重量分布与重心位置的图像索引方法[J]. , 2004, 30(01): 24-.
YU Wanbo1,2,3, ZHAO Jing2,3, LI Shumin3, WEI Xiaopeng2. Image Index Method Based on Weight and Gravity Center[J]. , 2004, 30(01): 24-.