An Image Classification Method Based on Multiple Level Spatial Visual Dictionary Ensemble

LUO Hui-lan; GUO Min-jie; KONG Fan-sheng

doi:10.3969/j.issn.0372-2112.2015.04.009

您当前的位置：

首页 >

文章列表页 >

An Image Classification Method Based on Multiple Level Spatial Visual Dictionary Ensemble

更新时间：2025-07-16

- An Image Classification Method Based on Multiple Level Spatial Visual Dictionary Ensemble
- Acta Electronica Sinica Vol. 43, Issue 4, Pages: 684-693(2015)
- 作者机构：
  
  1. 江西理工大学信息工程学院,江西,赣州,341000
  2. 浙江大学计算机科学技术学院,浙江,杭州,310027
  3. 江西理工大学信息工程学院,江西,赣州,341000
  4. 浙江大学计算机科学技术学院,浙江,杭州,310027
- 作者简介：
- 基金信息：
- DOI：10.3969/j.issn.0372-2112.2015.04.009
  CLC： TP391
- Published：2015
- 稿件说明：
移动端阅览
LUO Hui-lan, GUO Min-jie, KONG Fan-sheng. An Image Classification Method Based on Multiple Level Spatial Visual Dictionary Ensemble[J]. Acta Electronica Sinica, 2015, 43(4): 684-693.
DOI：

LUO Hui-lan, GUO Min-jie, KONG Fan-sheng. An Image Classification Method Based on Multiple Level Spatial Visual Dictionary Ensemble[J]. Acta Electronica Sinica, 2015, 43(4): 684-693. DOI： 10.3969/j.issn.0372-2112.2015.04.009.

摘要

针对单一特征时存在提取的信息量不足

对图像内容描述比较片面

提出将传统的SIFT特征与KDES-G特征进行串行融合

生成一个联合向量作为新的特征向量.针对传统的视觉词典构造方法缺乏考虑视觉词汇在空间的分布特点

本文引入图像空间信息

提出了一种空间视觉词典的构造方法

先对图像进行空间金字塔划分

再把空间各子区域内的特征分别聚类

构建属于对应子空间区域的空间视觉词典.在图像表示阶段

图像各子区域内的特征基于其对应的空间视觉词典进行LLC稀疏编码

根据各子区域对图像贡献程度的不同

把编码后各子区域的特征向量赋予不同的权重加权处理

再连接形成最终的图像描述.最后

利用线性SVM进行图像分类

实验结果表明了本文方法的有效性和鲁棒性.

Abstract

Using single feature of images to describe the image content is one-sided because of insufficient information.To solve this problem

combing SIFT and KDES-G features to describe images is proposed by generating a joint vector as a new feature vector.Considering images' spatial information

an image classification method based on the spatial visual dictionary is proposed.Images are first divided into sub-regions according to spatial pyramid

and the spatial visual dictionaries are respectively constructed by grouping features of each region into a number of clusters.The features of each region are coded by LLC based on its corresponding dictionary

and then the coded feature vectors are given different weights according to the different contribution of each region.After that

the feature vectors of different regions are concatenated and regarded as the final image description.Finally

a linear SVM is used to classify images.Experimental results show that the proposed method has better performance and robustness compared with some state-of-the-art works.

关键词

Keywords

references

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Variable Horizon Multi-Directional Scanning Method for Time Series Anomaly Detection

Image Classification Network of Background Perception Mechanism

Low-Light Small Target Detection Method Combining Feature Fusion Enhancement and Detail Features

Object Detection Based on EIMYOLO for High-Resolution Remote Sensing Images

Research on Improved Localization Based on Real-Time Semantic Chain List Systems

Related Author

HUANG Yu-zhe

GUAN Yong-yuan

WEI Song-jie

YUAN Heng

RAN Chao

ZHANG Sheng-chong

JIANG Ze-tao

CHENG Liu-ming

Related Institution

School of Computer Science and Engineering, School of Cyber Science and Engineering, Nanjing University of Science and Technology

School of Software, Liaoning Technical University

Key Laboratory of Optoelectronic Information Control and Security Technology

Guangxi Key Laboratory of Image and Graphics Processing and Intelligent Processing, Guilin University of Electronic Technology

School of Information and Technology, Shanxi University

⁰