您的位置:首页  > 论文页面

基于空间加权特征包模型的图像检索

发表时间:2013-02-28  浏览量:1661  下载量:491
全部作者: 王传乾,张宝昌
作者单位: 北京航空航天大学自动化科学与电气工程学院飞行器控制一体化技术重点实验室
摘 要: 提出2种全新的空间加权方法,一种方法对图像的子块计算局部方差作为子块内特征点的权重,另一种方法通过计算图像子块之间的相邻关系获得图像子块内特征点的权重。最后根据图像背景特征,提出了级联检索策略。研究提出的方法简单,执行效率高,而且相比无序的“特征包”(bag of features,BoF)方法,检索准确率有了显著提高。
关 键 词: 人工智能;空间加权;特征包;级联检索
Title: Image retrieval based on spatial weighting bag of features
Author: WANG Chuanqian, ZHANG Baochang
Organization: Science and Technology on Aircraft Control Laboratory, School of Automation Science and Electrical Engineering, Beihang University
Abstract: Two novel spatial weighting approaches were proposed. One method was based on local variance of blocks of an image, and the other was based on the adjacent relation of the blocks. After analyzing the background properties of images, we developed a cascaded retrieval approach. Methods proposed in this paper are simple and effective, and when comparing to the traditional orderless bag of features (BoF) model, these methods improve the accuracy rate of retrieval significantly.
Key words: artificial intelligence; spatial weighting; bag of features; cascaded retrieval
发表期数: 2013年2月第4期
引用格式: 王传乾,张宝昌. 基于空间加权特征包模型的图像检索[J]. 中国科技论文在线精品论文,2013,6(4):296-301.
 
0 评论数 0
暂无评论
友情链接