光学与光电技术, 2017, 15 (6): 24, 网络出版: 2018-01-09
基于Gabor特征多分类器融合的植物叶片识别方法
Research on Leaf Recognition Based on Gabor Feature and Multi-Classifier Fusion
植物叶片识别 植物分类 Gabor特征 D-S证据理论 多分类器 leaf recognition plants classification Gabor characteristic D-S evidence theory mulit-classifier
摘要
叶片识别是植物分类领域中最为重要的技术环节。以往的识别研究以单一的分类器进行分类识别,在识别过程中存在局限性,降低了植物分类的正确率。提出了一种基于D-S证据理论的多分类器融合识别方法。该方法利用多个分类器的识别结果构造D-S分配函数,通过D-S融合输出最终结果,能够在一定程度提高植物叶片的识别正确率。实验结果表明,该方法相对于单分类器可以取得更好的识别效果。
Abstract
Leaf recognition is the most important part in plants classification. However, previous studies only use single classifier to do recognition, the accuracy of the classification is low since the disadvantage of single classification method. To improve the classification accuracy, multi-classifier fusion using D-S evidence theory is proposed in this paper. This method gets the D-S distribution function using the results from mulit-classifiers. The classification accuracy can be improved based on this fusion result. The experimental results show that the proposed method works well, and it can improve the accuracy of leaf recognition.
陈筱勇, 马兰. 基于Gabor特征多分类器融合的植物叶片识别方法[J]. 光学与光电技术, 2017, 15(6): 24. CHEN Xiao-yong, MA Lan. Research on Leaf Recognition Based on Gabor Feature and Multi-Classifier Fusion[J]. OPTICS & OPTOELECTRONIC TECHNOLOGY, 2017, 15(6): 24.