当前位置>主页 > 期刊在线 > 信息技术 >

信息技术22年20期

基于 Google Earth Engine 的福州市 耕地信息提取研究
徐晓臣,李昺星
(中水北方勘测设计研究有限责任公司,天津 300222)

摘  要:南方山地丘陵地区耕地资源稀缺,耕地分布较为分散,导致现有的土地利用产品存在精度较低的情况。针对上述问题,文章以福州市为例,基于 GEE 云计算平台,借助 Sentinel-1 与 Sentinel-2 数据,探究不同特征组合下随机森林耕地信息的提取效果。结果表明,光学影像 Sentinel-2 数据的分类精度高于 Sentinel-1 数据,光谱指数和地形特征的引入使得分类精度得到明显提升;通过结合 Sentinel-1 和 Sentinel-2 数据构建最佳特征组合,总体精度达到 92.37%,Kappa 系数为 0.89。总体而言,基于 GEE 云平台的 Sentinel-1 和 Sentinel-2 数据适用于南方多云多雨复杂地形地区的耕地信息提取。


关键词:GEE;耕地;Sentinel-1;Sentinel-2;随机森林



DOI:10.19850/j.cnki.2096-4706.2022.20.003


中图分类号:TP181                                       文献标识码:A                                      文章编号:2096-4706(2022)20-0011-04


Research on the Extraction of Cultivated Land Information in Fuzhou City Based on Google Earth Engine

XU Xiaochen, LI Bingxing

(China Water Resources Beifang Investigation, Design and Research Co., Ltd., Tianjin 300222, China)

Abstract: The scarcity of cultivated land resources and the dispersion of cultivated land distribution in mountainous and hilly areas in the south lead to the low accuracy of existing land using products. To solve the above problems, this paper takes Fuzhou City as an example, based on the GEE cloud computing platform, with the help of Sentinel-1 and Sentinel-2 data, to explore the effect of random forest cultivated land information extraction under different feature combinations. The results show that the classification accuracy of optical image Sentinel-2 data is higher than that of Sentinel-1 data, and the introduction of spectral index and terrain features makes the classification accuracy significantly improved; by combining Sentinel-1 and Sentinel-2 data to build the best feature combination, the overall accuracy reaches 92.37%, and the Kappa coefficient is 0.89. In general, the Sentinel-1 and Sentinel-2 data based on the GEE cloud platform are applicable to the extraction of cultivated land information in the southern cloudy and rainy complex terrain area.

Keywords: GEE; cultivated land; Sentinel-1; Sentinel-2; random forest


参考文献:

[1] THENKABAIL P S. Global Croplands and their Importance for Water and Food Security in the Twenty-first Century: Towards an Ever Green Revolution that Combines a Second Green Revolution with a Blue Revolution [J].Remote Sensing,2010,2(9):2305-2312.

[2] 柏亦琳,饶夏晴,蔡岱桓,等 . 基于分水岭算法和混合效应模型的小麦图像分割——以兖州市王因镇,黄屯镇为例 [J]. 自动化应用,2021(4)166-170.

[3] 周楠,杨鹏,魏春山,等 . 地块尺度的山区耕地精准提取方法 [J]. 农业工程学报,2021,37(19):260-266.

[4] 毛丽君,李明诗 .GEE 环境下联合 Sentinel 主被动遥感数据的国家公园土地覆盖分类 [J/OL]. 武汉大学学报 ( 信息科学版 ),

2021:1-19[2022-08-01].https://kns.cnki.net/kcms/detail/detail.aspx?dbcode=CAPJ&dbname=CAPJLAST&filename=WH CH20210322000&uniplatform=NZKPT&v=kv866gsLbeOMvLWbj tPhyQ5ScwMDYYbChXGAzM-satjhW2k-su0d6VwdPQazFbT7.

[5] 林娜,王伟,王斌 . 基于随机森林和 Landsat8 OLI 影像的脐橙果园种植信息提取 [J]. 地理空间信息,2021,19(11):96-100+8-9.


作者简介:徐晓臣(1990—),男,汉族,山东曲阜人,工程师,硕士研究生,研究方向:测绘工程、摄影测量与遥感。