计算机应用 ›› 2013, Vol. 33 ›› Issue (08): 2310-2312.

• 多媒体处理技术 • 上一篇    下一篇

基于潜在狄利克雷分配的图像多层视觉表示方法

李冬睿1,李梅2   

  1. 1. 广东农工商职业技术学院 计算机系,广州 510507;
    2. 广东农工商职业技术学院 网络中心,广州 510507
  • 收稿日期:2013-03-07 修回日期:2013-04-23 出版日期:2013-08-01 发布日期:2013-09-11
  • 通讯作者: 李冬睿
  • 作者简介:李冬睿(1983-),男,广东广州人,讲师,硕士,主要研究方向:图形图像处理、模式识别、嵌入式控制;
    李梅(1973-),女,山西太原人,副教授,硕士,主要研究方向:网络编程、人工智能。
  • 基金资助:
    广东省科技厅资助项目;广东省经济和信息化委员会资助项目

Image multilayer visual representation method based on latent dirichlet allocation

LI Dongrui1,LI Mei2   

  1. 1. Department of Computer, Guangdong Agriculture Industry Business Polytechnic College, Guangzhou Guangdong 510507, China
    2. Network Center, Guangdong Agriculture Industry Business Polytechnic College, Guangzhou Guangdong 510507, China
  • Received:2013-03-07 Revised:2013-04-23 Online:2013-09-11 Published:2013-08-01
  • Contact: LI Dongrui

摘要: 针对前馈型图像多层视觉表示方法难以处理局部模糊情况,提出一种基于潜在狄利克雷分配(LDA)的图像多层视觉表示方法——LDA-IMVR。通过递归的概率分解方式,获得LDA的递归生成模型;同时,通过学习和推断多层结构的所有分层,以及利用反馈方式来提高分类学习性能。在Caltech 101数据集上的实验结果表明,与相关的多层视觉表示方法比较,LDA-IMVR提高了数据对象的分类性能,并且在分量学习和图像特征区域可视化方面也得到了较好的效果。

关键词: 分层视觉表示, 计算机视觉, 潜在狄利克雷分配, 递归, 反馈

Abstract: Image layer visual representation has been currently used in computer vision field, but it is difficult for feed-forward image multilayer visual representation methods to deal with local ambiguities. An image multilayer visual representation method based on Latent Dirichlet Allocation (LDA) named LDA-IMVR was proposed. It derived a recursive generative model of LDA by implementing recursive probabilistic decomposition process. Meanwhile, it learned and deduced all layers of the hierarchy together, and improved classification and learning performance by using feed-back style. The approach was tested on Caltech 101 dataset. The experimental results show that the proposed method improves classification performance of objects compared with related hierarchical approaches, and it achieves better results in learned components and image patches visualization.

Key words: layer visual representation, computer vision, Latent Dirichlet Allocation (LDA), recursive, feed-back

中图分类号: