Research work · 2009
Regional Classification According to Appearance Character of Domestic Flue-cured Tobacco Based on Two Dimensional Graph Theory Cluster Analysis
Wei Chun-Yang, Xinmin Wang, Cheng Sen, Cai Xian-Jie, Song Ji-Zhen, Yin Qi-Sheng
Tobacco Science & Technology
Abstract
Based on the appearance quality data of cutters of domestic flue-cured tobacco leaf from 2003 to 2008,the tobacco's appearance character was regionally classified via two dimensional graph theory cluster analysis.The results showed that:1)the regional difference of appearance character was mainly represented by leaf's color intensity,oil and body;2)taking the similarities of appearance character and eco-environment,and geographical neighborhood into account,appearance character was divided into 5 regions and 20 clusters.Cluster 1-4 was classified as Region Ⅰ,including Heilongjiang,Jilin,Liaoning and Gansu;Cluster 5-8 as Region Ⅱ,including Shandong,Henan,north of Sha'anxi and north of Anhui;Cluster 9-10 as Region Ⅲ,including Chongqing,Hubei,south of Sha'anxi,west of Hunan and northwest of Guizhou;Cluster 11-16 as Region Ⅳ,including Guangdong,Jiangxi,Fujian,south of Anhui,southeast of Hunan and northeast of Guangxi;Cluster 17-20 as Region Ⅴ,including Yunnan,middle and south of Guizhou,south of Sichuan and southwest of Guangxi.