基于国产高时空分辨率卫星影像的作物种植信息提取研究

发布时间：2018-05-31 01:24

本文选题：种植结构 + 遥感影像　；参考：《福建农业学报》2017年05期

【摘要】：在面向对象技术支持下,首先利用高空间分辨率ZY-3遥感影像提取农田地块专题信息;然后在地块边界控制下以地块对象为单元融入HJ-1及GF-1中分传感器的多时相光谱信息,获取作物生长关键期内的时间序列光谱特征;最后,结合不同作物的物候差异性规律构建作物种植信息提取模型,对甘蔗和水稻进行识别。结果表明,所有地类的总体分类精度为86.80%,Kappa系数为0.84,总体分类效果良好。甘蔗的制图精度和用户精度分别达到92.11%和90.91%,水稻的制图精度和用户精度分别达到88.89%和90.91%。说明协同利用国产卫星的高空间和高时间分辨率影像数据提取作物种植信息确实可行,可作为作物种植面积和种植结构的精细化、快速调查方法。
[Abstract]:With the support of object oriented technology, the special information of farmland block is extracted by high spatial resolution ZY-3 remote sensing image, and then the multi-phase spectral information of sub-sensor in HJ-1 and GF-1 is integrated into the block object unit under the control of block boundary. The spectral characteristics of the time series in the critical period of crop growth were obtained. Finally, combining with the phenological differences of different crops, the information extraction model of crop planting was constructed to identify sugarcane and rice. The results show that the overall classification accuracy of all ground classes is 86.80 and the Kappa coefficient is 0.84, and the overall classification effect is good. The mapping accuracy and user accuracy of sugarcane were 92.11% and 90.91% respectively, and the mapping accuracy and user precision of rice reached 88.89% and 90.91% respectively. The results show that it is feasible to extract crop planting information by using the high spatial and high time resolution image data of domestic satellites, which can be used as a fine and rapid investigation method for crop planting area and structure.
【作者单位】：广西农业科学院农业科技信息研究所;中国科学院遥感与数字地球研究所;中国科学院大学;
【基金】：广西科学研究与技术开发计划项目(桂科合14125008-1-6)
【分类号】：S127

【相似文献】