摘要
为探索基于卫星平台的油菜菌核病遥感监测方法,于2011-2012年在新疆伊犁第四师77团油菜生长季节,对选定的健康、病害油菜地块叶面积指数(LAI)、花面积指数(FAI)和角果面积指数(PAI)及光谱指数进行了研究。结果显示:健康油菜与病害油菜的归一化植被指数(NDVI)、比值植被指数(RVI)和差值植被指数(DVI)在发病前无明显著差异,但在发病之后有显著或极显著差异,基于显著差异性光谱指数采用的4种病害与非病害油菜分类方法中,以最大似然法分类结果与田间调查结果最接近,其kappa系数为0.72,达到高度的一致性。结论:利用NDVI、RVI和DVI可区分病害油菜与健康油菜,最大似然法是病害油菜与健康油菜分类的最佳方法,利用环境卫星多光谱数据进行油菜病害监测是可行的,精度符合生产要求。
In order to explore the remote sensing monitoring method for selerotinia based on the satellite platform, we carried out the comparative analysis in rape growing season of 2011-2012, in 77th farm, Ill, Xinjiang. We selected healthy and diseased rape plots for the study of leaf area index（LAD, the flower area index（FAD, the pod area index（PAD and spectral index. The results showed that,NDV1,RVI and DVI of the healthy rape and diseased rape didn＇t have significant difference before the onset, but after the onset, the difference became significant or very significant. Among the four classification methods based on significant differences in spectral index,the maximum likelihood method was the closest to the field investigation results, with the kappa coefficient 0.72, reaching a high degree of consistency （substantial）. Conclusion： NDVI, RVI and DVI can be used to classify the healthy and diseased rape,the maximum likelihood method is the best one for the healthy and diseased rape classification, and the environmental satellite multispeetral data is feasible in rape diseases monitoring.
出处
《石河子大学学报（自然科学版）》
CAS
2013年第5期568-574,共7页
Journal of Shihezi University(Natural Science)
基金
国家科技支撑计划项目(2011BAD16B14)
国家自然科学基金项目(31071371
41161068)
关键词
油菜
光谱植被指数
农学参数
生育期
rapeseed, spectral index, vegetation parameters, growth period
