基于层次化特征融合的花卉图像分类
首发时间:2014-10-07
摘要:针对花卉类间相似和类内差异的问题,传统的特征融合只是简单的把多个特征融合在一起,忽略了不同特征对不同的花卉类别的影响。为解决这一问题,本文提出一种基于层次化特征融合的分类方法。该方法针对每类花卉的颜色、形状特征进行单独训练,得到特征对应于类别的得分值,能够对表现差异大的花卉进行有效的分类,同时又能够克服类内变化大的问题。实验结果表明,该方法比直接的特征融合方式分类效果更好,能够有效地分类出类间相似度比较高以及类内差异性比较大的花卉。
关键词: 层次化特征融合 花卉图像分类; 精细图像
For information in English, please click here
Hierarchical Feature Fusion based Flower Image Classification
Abstract:Due to the inter-class similarity and intra-class difference issue, the traditional feature fusion methods are only the linear combination of each feature and neglect the influence of different features for different flower categories. To solve that problem, we propose a hierarchical feature fusion based classification method. In this method, we train a classifier using color and shape feature for each type individually and obtain score of each category. This method is able to overcome the intra-class difference problem. Experiment results show that our proposed hierarchical feature fusion based classification method superior to some traditional classical feature fusion methods and can classify the flower with high inter-class similarity and intra-class difference.
Keywords: Hierarchical Feature Fusion Flower Image Classification Fine-grained Image
基金:
论文图表:
引用
No.4610754981334141****
同行评议
共计0人参与
勘误表
基于层次化特征融合的花卉图像分类
评论
全部评论0/1000