计算机应用 ›› 2015, Vol. 35 ›› Issue (9): 2629-2635.DOI: 10.11772/j.issn.1001-9081.2015.09.2629

• 虚拟现实与数字媒体 • 上一篇    下一篇

基于视觉显著性检测的图像分类方法

刘尚旺1,2, 李名1,2, 胡剑兰1,2, 崔艳萌1,2   

  1. 1. 河南师范大学 计算机与信息工程学院, 河南 新乡 453007;
    2. "智慧商务与物联网技术"河南省工程实验室(河南师范大学), 河南 新乡 453007
  • 收稿日期:2015-04-27 修回日期:2015-06-21 出版日期:2015-09-10 发布日期:2015-09-17
  • 通讯作者: 刘尚旺(1973-),男,河南新乡人,副教授,博士,主要研究方向:生物图像处理,shwl2012@hotmail.com
  • 作者简介:李名(1981-),男,河南新乡人,副教授,博士,主要研究方向:图像信息隐藏;胡剑兰(1991-),女,河南信阳人,硕士研究生,主要研究方向:生物图像处理;崔艳萌(1989-),女,河南商丘人,硕士研究生,主要研究方向:生物图像处理。
  • 基金资助:
    国家自然科学基金资助项目(U1304607);河南省高等学校重点项目(15A520080,15A520020);河南师范大学博士科研启动基金资助项目(qd12138,qd14134)。

Image classification method based on visual saliency detection

LIU Shangwang1,2, LI Ming1,2, HU Jianlan1,2, CUI Yanmeng1,2   

  1. 1. College of Computer and Information Engineering, Henan Normal University, Xinxiang Henan 453007, China;
    2. Henan Engineering Laboratory of Intelligence Business and Internet of Things (Henan Normal University), Xinxiang Henan 453007, China
  • Received:2015-04-27 Revised:2015-06-21 Online:2015-09-10 Published:2015-09-17

摘要: 针对传统的图像分类方法对整个图像不分等级处理以及缺乏高层认知的问题,提出了一种基于显著性检测的图像分类方法。首先,利用视觉注意模型进行显著性检测,得到图像的显著区域;然后,利用Gabor滤波方法和脉冲耦合神经网络模型,分别提取该显著区域的纹理特征和时间签名特征;最后,根据提取的纹理特征和时间签名特征,利用支持向量机实现图像分类。实验结果表明,所提方法在SIMPLIcity图像数据集上平均分类正确率达到94.26%,在Caltech数据集上平均分类正确率为95.43%,从而证明,显著性检测与有效的特征提取对图像分类有重要影响。

关键词: 视觉注意模型, 显著区域, 脉冲耦合神经网络, Gabor滤波, 图像分类

Abstract: To solve the problem that traditional image classification methods deal with the whole image in a non-hierarchical way, an image classification method based on visual saliency detection was proposed. Firstly, the visual attention model was employed to generate the salient region. Secondly, the texture feature and time signature feature of the image were extracted by Gabor filter and pulse coupled neural network, respectively. Finally, the support vector machine was adopted to accomplish image classification according to the features of the salient region. The experimental results show that the image classification precision rates of the proposed method in SIMPLIcity and Caltech are 94.26% and 95.43%, respectively. Obviously, saliency detection and efficient image feature extraction are significant to image classification.

Key words: visual attention model, salient region, Pulse Coupled Neural Network (PCNN), Gabor filter, image classification

中图分类号: