Computer Engineering and Applications ›› 2018, Vol. 54 ›› Issue (8): 178-182.DOI: 10.3778/j.issn.1002-8331.1610-0018

Previous Articles     Next Articles

Image classification based on RGB-D fusion feature

XIANG Chengyu, WANG Dongli, ZHOU Yan, LI Yafang   

  1. School of Information Engineering, Xiangtan University, Xiangtan, Hunan 411100, China
  • Online:2018-04-15 Published:2018-05-02

基于RGB-D融合特征的图像分类

向程谕,王冬丽,周  彦,李雅芳   

  1. 湘潭大学 信息工程学院,湖南 湘潭 411100

Abstract: The classic image classification algorithms are mostly based on RGB or grayscale images, and the depth information of the object or scene has not been utilized?effectively. To solve this problem, this paper proposes an image classification method based on RGB-D fusion feature. Firstly, the dense SIFT feature of color image is fused with the global Gist feature of the depth image to generate a combined vector. Secondly, the improved K-means algorithm is used to build the visual dictionary of the fusion feature, overcoming the dependence on the initial point selection of traditional K-means algorithm. Moreover, in the stage of image representation, the approximate LLC feature coding method is introduced to operate sparse coding on feature base and its corresponding visual dictionary. Finally, the linear SVM is used for image classification. The experimental results show that the proposed algorithm can effectively improve the classification accuracy.

Key words: depth image, dense Scale Invariant Feature Transform(SIFT) feature, Gist feature, K-means, Locality-constrained Linear Coding(LLC) sparse coding

摘要: 当前经典的图像分类算法大多是基于RGB图像或灰度图像,并没有很好地利用物体或场景的深度信息,针对这个问题,提出了一种基于RGB-D融合特征的图像分类方法。首先,分别提取RGB图像dense SIFT局部特征与深度图Gist全局特征,然后将得到的两种图像特征进行特征融合;其次,使用改进K-means算法对融合特征建立视觉词典,克服了传统K-means算法过度依赖初始点选择的问题,并在图像表示阶段引入LLC稀疏编码对融合特征与其对应的视觉词典进行稀疏编码;最后,利用线性SVM进行图像分类。实验结果表明,所提出的算法能有效地提高图像分类的精度。

关键词: 深度图像, dense尺度不变特征变化(SIFT)特征, Gist特征, K-means算法, 局部约束线性编码(LLC)稀疏编码