摘要
为准确快速获取夏季玉米四叶期、拔节期、抽穗期和花粒期的植被覆盖度信息,利用无人机获取玉米田间可见光图像,对图像可见光波段提取的多种植被指数进行分析和比较,选择差异植被指数(Visible-band difference vegetation index,VDVI)、过绿指数(Excess green,EXG)和归一化绿蓝差异指数(Normalized green-blue difference index,NGBDI),结合监督分类提取了玉米4个时期的植被覆盖度信息。通过对试验田4个阶段的单幅图像监督分类处理,将其目标物分为土壤和玉米植被两类;分别统计监督分类后图像中土壤和玉米的VDVI像元直方图,将两者的像元直方图交点作为植被覆盖度提取阈值,同理获得EXG和NGBDI对应的玉米植被覆盖度提取阈值;利用获取的玉米植被3种覆盖度提取阈值,对玉米4个时期的植被覆盖度进行提取,并对提取精度进行了验证。结果表明,VDVI对应4个生长时期的植被覆盖度提取误差分别为1. 21%、4. 88%、2. 31%和3. 61%; EXG对应的植被覆盖度提取误差分别为1. 38%、1. 25%、0. 89%和0. 33%; NGBDI提取误差为1. 61%、3. 31%、1. 99%和3. 25%,EXG在夏季玉米4个生长时期的植被覆盖度提取效果最好。将玉米4个生长时期单幅图像确定的阈值作为固定阈值,对剔除确定阈值的单幅图像的试验田全景图像进行植被覆盖度提取,并对提取效果进行验证。结果表明,采用监督分类与可见光植被指数统计直方图相结合确定阈值的方法提取玉米植被覆盖度效果较好。
In order to accurately and rapidly obtain the vegetation coverage information of summer corn during the stages of four-leaf, jointing, heading and flowering, unmanned aerial vehicles (UAV) was used to obtain visible light images of corn field, and various vegetation indices extracted from visible light bands were analyzed and compared. Visible-band difference vegetation index (VDVI), excess green (EXG) and normalized green-blue difference index (NGBDI) were used to extract the corn vegetation coverage information of the four stages combined with supervised classification method. In the research process, targets in a single image of the experimental field were divided into soil and corn vegetation in the four stages of the corn. The VDVI pixel histograms of soil and corn classified by supervised classification method were counted respectively, and the intersection points of pixel histogram were used as the threshold of vegetation coverage extraction. Similarly, the threshold of corn vegetation coverage extraction corresponding to EXG and NGBDI was obtained. Finally, the corn vegetation coverage of the four stages was extracted by the three extraction thresholds. The errors of vegetation coverage extraction corresponding to the four growth stages of VDVI were 1.21%, 4.88%, 2.31% and 3.61%, respectively;EXG were 1.38%, 1.25%, 0.89% and 0.33%, respectively;and NGBDI were 1.61%, 3.31%, 1.99% and 3.25%, respectively. It was found that EXG had the best effect on vegetation coverage extraction during the four stages of corn. The value of threshold determined by the single image of the four corn growth stages was used as a fixed threshold, and the vegetation coverage was extracted from the panoramic image of the experimental field that had removed the single image which was used as determining threshold value, and the extraction effect was verified. The results showed that the variation of extraction error was small, indicating that the method using the supervised classification combined with the statistical histogram of visible vegetation index to determine the threshold value was better.
作者
赵静
杨焕波
兰玉彬
鲁力群
贾鹏
李志铭
ZHAO Jing;YANG Huanbo;LAN Yubin;LU Liqun;JIA Peng;LI Zhiming(School of Agricultural Engineering and Food Science,Shandong University of Technology,Zibo 255049,China;International Precision Agriculture Aviation Application Technology Research Center,Shandong University of Technology,Zibo 255049,China;School of Transportation and Vehicle Engineering,Shandong University of Technology,Zibo 255049,China)
出处
《农业机械学报》
EI
CAS
CSCD
北大核心
2019年第5期232-240,共9页
Transactions of the Chinese Society for Agricultural Machinery
基金
山东省"引进顶尖人才‘一事一议’专项经费"项目
中央引导地方科技发展专项资金项目
关键词
夏季玉米
无人机可见光图像
植被覆盖度
植被指数
阈值提取
summer corn
visible light images of UAV
vegetation coverage
vegetation index
threshold extraction
