结合HJ卫星影像和最小二乘孪生支持向量机的小麦蚜虫遥感监测

发布时间：2018-11-27 07:28

【摘要】：为了准确、及时地监测小麦蚜虫发生情况,利用野外定位调查数据及环境与灾害监测预报小卫星星座HJCCD和HJ-IRS影像数据,在北京市通州区和顺义区小麦蚜虫发生的关键生育期(灌浆期),提取对蚜虫病情影响较大的小麦长势因子和生境因子,利用最小二乘孪生支持向量机建立该研究区的小麦蚜虫监测模型,并与传统支持向量机、费歇尔线性判别分析和学习矢量量化神经网络模型的监测结果进行对比。结果表明:最小二乘孪生支持向量机模型的总体监测精度达到86.4%,优于传统支持向量机模型(77.3%)、费歇尔线性判别分析模型(77.3%)和学习矢量量化神经网络模型(72.7%),取得了较好的监测效果。
[Abstract]:In order to accurately and timely monitor the occurrence of wheat aphids, the field location survey data and the small satellite constellation HJCCD and HJ-IRS image data of environment and disaster monitoring and forecasting were used. Wheat growth factors and habitat factors, which have great influence on aphid disease, were extracted at the key growth stage (grain filling stage) of wheat aphids in Tongzhou and Shunyi districts of Beijing. The monitoring model of wheat aphid in this study area was established by using least square twin support vector machine (LS-SVM). The results were compared with those of traditional SVM, Fischer linear discriminant analysis and learning vector quantization neural network model. The results show that the overall monitoring accuracy of the least square twin support vector machine model is 86.4, which is better than that of the traditional support vector machine model (77.3%). Fisher linear discriminant analysis model (77.3%) and learning vector quantization neural network model (72.7%) have achieved better monitoring results.
【作者单位】：安徽大学安徽省农业生态大数据工程实验室;安徽大学电子信息工程学院;中国科学院遥感与数字地球研究所数字地球重点实验室;中国科学院南京地理与湖泊研究所;
【基金】：国家自然科学基金(61672032,41271412) 安徽省自然科学基金(1408085MF121,1608085MF139) 安徽省科技计划项目(16030701091,1604A0702016) 中国科学院国际合作局对外合作重点项目(131211KYSB20150034)
【分类号】：S435.122.2;TP18

【相似文献】