基于支持向量机的植物叶分类识别研究
首发时间:2015-11-26
摘要:该研究持有初次分类与再次分类的思想,先使用叶图像的形状特征与纹理特征进行植物种类的初次分类;再使用该研究提出的叶脉特征,由叶脉的分叉点和叶脉端点的数量和各自到中心点距离线性归一化后得出的叶脉特征对未识别出的、形状相似度高的植物叶图像实行再次分类。该研究对100种植物叶片分类,初次分类识别率为94.1%,再次分类后正确识别率为96.5%。
关键词: 植物叶分类;叶脉分叉点;叶脉端点;初次分类;再次分类
For information in English, please click here
Leaf recognition for plant based on Support Vector Machine
Abstract:The allocation and re-allocation of the initial study of thought holds. Firstly, the initial classification of plant species use leaf shape and texture feature image to classify; then use vein characteristics of the study proposed to re-classify the un-identified plant leaf images. The study of 100 kinds of plant leaves classification, the initial classification was 94.1%, after once again classified correctly identify 96.5%.
Keywords: leaf images classification veins endpoint first classification second classification veins branching point
基金:
论文图表:
引用
No.4661939105892914****
同行评议
共计0人参与
勘误表
基于支持向量机的植物叶分类识别研究
评论
全部评论0/1000