| 摘要: |
| 以农林业主要蛾类害虫为研究对象,应用其翅的数学形态特征在总科阶元上进行分类鉴定。对鳞翅目5总科40种蛾类的右前翅和右后翅的偏心率、球状性、叶状性、似圆度、矩形度、延长度以及不变矩Hu1、Hu2等共计26项与大小和方向均无关的数学形态特征,利用方差分析、逐步判别分析和聚类分析等方法论证各项数学形态特征在总科阶元上进行分类的可行性、可靠性和重要性,并且从数学形态学角度对同阶元昆虫类群的亲缘关系做了描述。结果表明:在总科阶元上可筛选出6个形态特征作为分类变量,它们的作用大小依次为:(FW矩形度、FWHu5、HW偏心率)>HW似圆度>(HW矩形度、HWHu5),回归判别和交叉判别的结果其正确率分别为100%和97.5%。 |
| 关键词: 蛾 形态特征 数字鉴定 亲缘关系 总科级阶元 |
| DOI：10.11841/j.issn.1007-4333.2013.04.15 |
| 投稿时间：2012-10-25 |
| 基金项目:河北省自然科学基金项目(C2012204008); 国家948项目(2013-4-75) |
|
| Systematic research on the classification of moths using math-morphological characters of wings Ⅰ：Superfamily level |
|
CAI Xiao-na1,2, HUANG Da-zhuang1, SHEN Zui-rui2, WANG Zhi-gang1, GAO Ling-wang2
|
| (1.College of Forestey, Agricultural University of Hebei, Baoding 071000, China;2.College of Agronomy and Biotechnology, China Agricultural University, Beijing 100193, China) |
| Abstract: |
| This paper is concerned on the superfamily level classification of moths which are harmful to agriculture and forestry,by using math-morphological characters(MMCs) of the wings.Twenty six MMCs were selected for being invariant to the image size and direction,such as eccentricity,sphericity,lobation,roundness,rectangularity,elongation and movement invariants including Hu1 and Hu2.These MMCs were extracted from the images of right forewing and right hindwing of forty moths (five superfamilies).By using the methods,including variance analysis,stepwise discriminant analysis and cluster analysis,each MMC was evaluated on the feasibility,reliability and importance in classification of the moths at superfamily level.The relationships among the moths at superfamily level were described from the perspective of mathematical morphology.The analytic results showed that six MMCs could be selected as the classification variables at superfamily level.The contribution of these variables were ranked as followings of (FW-rectangularity,FW-Hu5,HW-eccentricity)>HW-roundness>(HW-rectangularity,HW-Hu5).Accuracies resulting from the regression and the intersecting discriminant analysis were 100.0% and 97.5%,respectively. |
| Key words: moths morphological character digital identification phylogenetic relationship superfamily |
