摘要
植物的分类与识别是植物学研究和农林业生产经营中的重要基础工作,其主要依据是植物的外观特征。现代数量分类的方法就是通过大量提取特征数据,进行聚类分析,获得分类结果,并以此为根据进行植物的鉴别。传统做法都是手工测量采集原始数据,效率较低。由于外观特征可以通过数字图片方式获得,运用计算机图像处理分析等技术采集数据并做聚类分析将大大提高效率,关键问题在于特征的自动分析和获取。在基于叶子特征的计算机辅助植物识别模型和叶缘锯齿特征提取研究的基础上,提出了计算机辅助植物分类与识别的系统方案,并对相关技术进行了分析。
Classification and identification of plant are the groundwork for research on botany and management of forestry. Applying modern numerical taxonomy to plants is mostly based on the visible characteristics. The result is derived from the clustering analysis of many values of some visible characteristics. The traditional way of collecting the data manually is not so efficient. Since the visible characteristics are involved in the digital image of the plant or its local parts, image-processing technology can do the work automatically and efficiently. The focus here is how to obtain the characteristics automatically. On the basis of related researches such as computer-aided plant-identification model based on leaf characteristics and extracted sawtooth feature of leaf edge, a system scheme on plant classification and identification is developed. Then the feasibilities of technologies are also analyzed.
出处
《浙江林学院学报》
CAS
CSCD
北大核心
2004年第2期222-227,共6页
Journal of Zhejiang Forestry College
基金
浙江省教育厅资助项目(20020980)