Ramie Variety Identification Based on the Hyperspectral Parameters and the Stepwise Discriminant Analysis
CAO Xiao-lan1, 2, CHEN Xing-ming2, ZHANG Shuai2, CUI Guo-xian1*
1. Ramie Research Institute of Hunan Agricultural University, Changsha 410128, China
2. College of Information Science and Technology, Hunan Agricultural University, Changsha 410128, China
Abstract：The hyperspectral data on total 927 leaves of different genotypes, which come from 4 ramie varieties, were collected under the field cultivation conditions to explore the identification and classification of ramie varieties with the hyperspectral as the basis. According to the hyperspectral reflection curve of ramie leaves, two groups of feature parameters were extracted, namely, the hyperspectral wave-valley reflectance and position parameters(the group V1) as well as the skewness and kurtosis parameters(the group V2) . Then, by adopting the stepwise discriminant approach to screen different number of variables under different F-value settings, multiple Fisher linear discriminant functions based on these two groups of feature parameters were created respectively, and further, the created discriminant functions were comparatively analyzed from the computational complexity, the accuracy and the stability. So, we can come to the following conclusions: (1) For discriminant functions under all the combinations, the overall average accuracy was 91.1% and the overall standard deviation mean was 1.2%; (2) From the comprehensive trade-off perspective, when the number of variables was between 8 and 14, the discriminant effect of the group V2 was the best among all the combinations, namely, the computational complexity was in the middle level, and both the accuracy and the stability were over their corresponding average values; among them, the discriminant functions with 13 variables had the highest average accuracy and the lowest standard deviation, which were 94.2% and 0% respectively; (3) When the accuracy was considered preferentially and the number of variables changed between 15 and 22, the group V1 had the highest accuracy which was 95.5%, however, the computational complexity under this case was higher, the stability was in the middle level and the lowest standard deviation was 0.9%. Above results showed that it was feasible to utilize the hyperspectral parameters together with the stepwise discriminant approach.
曹晓兰，陈星明，张 帅，崔国贤. 高光谱参数和逐步判别的苎麻品种识别[J]. 光谱学与光谱分析, 2018, 38(05): 1547-1551.
CAO Xiao-lan, CHEN Xing-ming, ZHANG Shuai, CUI Guo-xian. Ramie Variety Identification Based on the Hyperspectral Parameters and the Stepwise Discriminant Analysis. SPECTROSCOPY AND SPECTRAL ANALYSIS, 2018, 38(05): 1547-1551.
[1] LI Jun(李 军). Agricultural Information Technology(农业信息技术). Beijing: Science Press(北京: 科学出版社)，2015. 98.
[2] LIU Yao, TAN Ke-zhu, CHEN Yue-hua, et al(刘 瑶，谭克竹, 陈月华，等). Soybean Science(大豆科学)，2016，35(4): 671.
[3] ZANG Zhuo, LIN Hui, YANG Min-hua(臧 卓, 林 辉, 杨敏华). Science of Surveying and Mapping(测绘科学)，2014，39(2)： 146.
[4] SHI Fei-fei, GAO Xiao-hong, YANG Ling-yu, et al(史飞飞, 高小红, 杨灵玉，等). Geography and Geo-Information Science(地理与地理信息科学)，2016，32(2): 32.
[5] WANG Dong, WU Jian(王 岽, 吴 见). Geography and Geo-Information Science(地理与地理信息科学)，2015，31(2): 29.
[6] CHAI Ying, RUAN Ren-zong, CHAI Guo-wu，et al(柴 颖, 阮仁宗, 柴国武，等). Remote Sensing for Land and Resource(国土资源遥感)，2016，28(3): 86.
[7] SU Jian-guang, DAI Zhi-gang(粟建光, 戴志刚). Chinese Hemp Plant Germplasm Resources and Their Main Characters(中国麻类作物种质资源及其主要性状). Beijing: China Agricultural Press(北京: 中国农业出版社)，2016. 9.
[8] JIA Jun-ping(贾俊平). Fundamental Statistic(统计学基础). Beijing: Renmin University of China Press(北京： 中国人民大学出版社)，2013. 89, 90.
[9] LU Wen-dai, ZHU Hong-bing(卢纹岱，朱红兵). SPSS Statistical Analysis(SPSS统计分析). Beijing: Publishing House of Electronics Indeustry(北京： 电子工业出版社)，2015，4: 493.
